Maximum Recycled Content Report 191009 - WRAP Recycled Content...glass. For whisky and gin, when considering the full sample, there was just one statistically significant difference

Final Report

Assessing the demand for glass containers with maximum recycled content

This project assessed the opportunity to manufacture glass containers from mixed colour recycled glass and ascertained the effect on consumer buying decisions for a number of products from the food, wine and spirits categories.

Project code: MSG029-005 ISBN: 1-84405-422-5 Research date: April 09 to August 09 Date: September 2009

WRAP helps individuals, businesses and local authorities to reduce waste and recycle more, making better use of resources and helping to tackle climate change. Document reference: [e.g. WRAP, 2006, Report Name (WRAP Project TYR009-19. Report prepared by…..Banbury, WRAP]

Written by: Nick Kirk, GTS and

Ed Sutherland, Steve Westerman, Peter Gardner and Liz Andrews, Psychology of Design Group, Institute of Psychological Sciences, University of Leeds.

Front cover photography: [Add description or title of image.] WRAP and Glass Technology Services Ltd believe the content of this report to be correct as at the date of writing. However, factors such as prices, levels of recycled content and regulatory requirements are subject to change and users of the report should check with their suppliers to confirm the current situation. In addition, care should be taken in using any of the cost information provided as it is based upon numerous project-specific assumptions (such as scale, location, tender context, etc.). The report does not claim to be exhaustive, nor does it claim to cover all relevant products and specifications available on the market. While steps have been taken to ensure accuracy, WRAP cannot accept responsibility or be held liable to any person for any loss or damage arising out of or in connection with this information being inaccurate, incomplete or misleading. It is the responsibility of the potential user of a material or product to consult with the supplier or manufacturer and ascertain whether a particular product will satisfy their specific requirements. The listing or featuring of a particular product or company does not constitute an endorsement by WRAP and WRAP cannot guarantee the performance of individual products or materials. This material is copyrighted. It may be reproduced free of charge subject to the material being accurate and not used in a misleading context. The source of the material must be identified and the copyright status acknowledged. This material must not be used to endorse or used to suggest WRAP’s endorsement of a commercial product or service. For more detail, please refer to WRAP’s Terms & Conditions on its web site: www.wrap.org.uk

Assessing the demand for glass containers with maximum recycled content 1

Executive Summary This study ‘Assessing the demand for glass containers with maximum recycled content’, was commissioned by WRAP (Waste & Resources Action Programme) and delivered by Glass Technology Services Ltd. One objective of this project was to assess the demand from retailers and brand owners for glass containers manufactured with maximised glass recycled content, remelted as mixed colour cullet. Additionally, the project sought to understand the marketability of such containers to consumers. In recent years the proportion of cullet collected as mixed colour as opposed to colour separated has increased significantly. This is in part due to a notable increase in domestic kerbside and hospitality sector collections, the majority of which collect the glass in a mixed colour format. This mixed cullet can be colour sorted at the processor to provide a segregated feedstock to the glass manufacturing industry. However, in some instances it may not be economic to run the colour sorting technology, as other markets such as export and aggregates are available and have cheaper processing costs and good revenues. The trend towards mixed colour glass collections has been accompanied by a decrease in traditional bottle bank collections which predominately collect glass as separate colours - green, amber and flint. If the growth in mixed colour glass collections continues, then less cullet colour separated at source will be available for remelt back into new glass containers. Remelt of cullet into new glass containers represents closed loop recycling and offers the best environmental option for the use of container cullet. In addition, there is currently an imbalance between the colours of glass that are recycled from the UK waste stream, and the colour demanded by UK container manufacturers. In 2008 51% of the glass recycled in the UK was green with flint making up 37.5% and amber at around 11.5%. The high volume of green glass in the UK waste stream is attributed to the large volumes of imported full green containers, particularly wine. UK container glass production consists of flint (64%), amber (20%) and green (16%) - with flint predominantly used in the food and spirit sectors. The result of this situation is that the UK has become a net importer of green glass, much of which is currently going into secondary markets such as aggregates. By challenging consumer expectation this work could stimulate a market within the container glass industry for the current surplus of green cullet. Two approaches to address the increasing proportion of cullet being collected as mixed colour are:

greater use of existing colour separation technologies. Glass processors already have in place technology to

optically sort mixed colour glass into separate colours. The yield from colour separation varies depending on

the quality of the input cullet and particle size, but is typically around 70%. However, the colour purity of

colour separated cullet does not match that of cullet collected as separate colours; and

increased manufacture of glass containers from colour mixed rather than single colour cullet.

This project addressed the latter possibility and the likely impact on consumers buying decisions on a range of products that are traditionally filled in flint. Replica containers with a green hue, representative of glass incorporating high levels of mixed colour cullet, were filled and the repackaged products were used to test consumer perceptions. For practical and cost reasons it was not viable to manufacture glass bottles and jars from mixed colour cullet specifically for this study. As a consequence, clear food and drink containers were coated with a transparent spray paint to simulate the colour of containers containing high recycled content as mixed colour cullet. The coated bottles and jars had their contents and label removed prior to spray coating and were then refilled and relabelled. In this study two spray colours were used:

“light” colour coating to represent bottles and jars manufactured using hospitality cullet; and

“dark” colour coating to represent bottles and jars manufactured using municipal cullet.

In practice, the difference between the “light” and “dark” colours was very slight. A range of products were covered by the study, including preserves, cook in sauces, table sauces, wine and spirits. For each product thirty samples from each product were taken, and their packaging treated as follows:


10 samples were kept in their original packaging (no coating);

10 were given a “light” colour coating to represent bottles and jars manufactured using hospitality cullet; and

10 were given a “dark” colour coating to represent bottles and jars manufactured using municipal cullet.

The two spray colours selected for this study were selected as reasonably representative of the typical outcome expected from manufacturing containers using mixed colour cullet. However in practice mixed colour cullet will vary in colour composition dependent on geographical region, glass collection method and also season. As such, if products with a high recycled content of mixed colour cullet were manufactured, variability in resulting glass colour would be seen. Consumer Perception Study To assess consumer perceptions of products with high mixed colour cullet recycled content, an independent study was conducted by the Psychology of Design Group at the University of Leeds. Two studies were commissioned for which a total of 116 participants were recruited:

Study 1, Participants = 79. In ‘Study 1’ participants were shown nine products, each presented in two

different glass colours: a green glass “dark” (containers manufactured using municipal cullet) and clear glass

(original uncoated product packaging); and

Study 2, Participants = 37. In Study Two an additional product was tested (giving 10 products in total), and

the comparison was made between flint glass (original uncoated product packaging as used in Study One)

and green glass “light” (containers manufactured using hospitality cullet).

Participants reflected the range of adult supermarket customers, insofar as they were recruited at a Sainsbury’s store and allotted to one of three age groups, 18-34, 35-55 and 55 and over; the male/female split in each group was approximately equal. A multi-methods approach was used to assess the influence of glass colour on consumers’ perception of the products:

a quantitative section of each testing session required participants to provide ratings of products on a variety

of different scales e.g. how appetising the product appeared and purchase likelihood, with the viewing

conditions representing a retail ‘on shelf’ setting. Participants were not informed of the manipulation of glass

colour at this stage; and

in a subsequent qualitative testing session participants contributed to a focus group discussion regarding the

products/containers and were made aware that the glass colour had been manipulated. Participants were

invited to consider the products in different contexts, including ‘in use’, and to facilitate discussions they were

shown part-filled examples of some containers. Participants were also invited to discuss issues relating to

recycling and how the ‘green’ products could be marketed.

The use of this multi-method approach enabled ‘triangulation’ of results, allowing confirmation and more detailed explanation of outcomes. Results For mayonnaise, the rating scale data and comments from the focus groups for both studies strongly indicated that the introduction of green glass would not be acceptable to consumers. Participants tended to attribute the colour of the container to properties of the product, and in the green containers (of both shades) the ratings of mayonnaise quality suffered and comments reflected a strong dislike (e.g., ‘I would never buy it in a million years’). For the remaining products the picture was less clear. A sample of the range of views is highlighted below:

For pasta sauce, ratings of purchase likelihood were lower when the product was presented in green glass in

Study One. However for jam, some ratings were more favourable when the product was presented in green

glass.

For whisky and gin, when considering the full sample, there was just one statistically significant difference in

ratings (gin was rated as purer in the flint bottle in Study Two). However, comments from the focus groups

indicated a tendency towards polarised views with some participants expressing a strong preference for the

flint bottle and some expressing a strong preference for the green bottle(s).


Consistent with this, there was a significant interactive effect indicating that the older participant group (55+

years) would be less likely to purchase whisky in a green bottle, whereas this was not the case for the other

age groups (Study One only).

For wine products there were no significant effects of glass colour, and comments suggest a good deal of

acceptance of the green bottles. This may be because green glass bottles are currently commonly seen for

these products. For several of the products tested interactive effects suggested that consumers’ perceptions

were dependent on demographic factors such as age and gender.

Across the range of products results indicated that differences in glass colour were often attributed to differences in the contents of the container, with comments indicating that products presented in green glass tended to be thought of as stronger/richer in taste, but less fresh/pure. Focus groups comments suggested that, for many participants, if sufficient explanation could be given as to why products were being sold in green rather than clear glass (such as the environmental implications), then the glass colour change would be acceptable. However, caution is needed in extrapolating from this reasoned response. The influence of non-conscious decision making is well known in the retail context. That is, consumers do not always make purchase decisions on the basis of reasoned arguments. Conclusions The influence of green-tinted glass containers on consumer perception is product dependent. For the majority of the products tested i.e. jam, wine, gin and whisky there was no statistically significant difference, and even a small advantage associated with the use of a green container (Basics jam in Study One). However, the study showed that for some products, such as mayonnaise, the use of green glass of any noticeable shade, is likely to prove unacceptable to consumers. It may be that the colour and/or nature of the product plays an important role in consumers’ perceptions and responses. For darker coloured food products and for alcoholic drink products the effects of container colour were less consistent. Ratings for purchase likelihood for pasta sauce were also higher when presented in flint glass (Study One). Caution must be exercised when interpreting these results for the following reasons. Firstly, results of the focus groups indicated that the change in container colour influenced participants’ perceptions of the products along a number of dimensions, including taste, freshness, and price. If consumers’ product preferences are a function of product attribute priorities and product attribute values, colour-related differences in outcome can be predicted. Secondly, there was a good deal of inter-individual variability in the nature of these effects. For some products similar average ratings for different glass colours seemed to mask more polarised inter-participant views, that is, greater standard deviation. However some of this spread was predictable in terms of the interactive effects of age or gender of participants. Thirdly, statistical tests are designed to identify the reliable presence of a difference between means rather than the absence of the same. If the study were repeated with a larger sample and/or different methods, statistically reliable differences may be identified. That is these results should be treated as indicative rather than absolute truths. Finally, the focus groups identified that other design features, such as label colour, label size, cap colour, and fill level, influence consumer perceptions of glass container colour. These factors should feature in the consideration of any container design changes. Retailer & Brand owner Demand The demand from retailers and brand owners for glass containers with high mixed colour glass recycled content will be ascertained by a survey in October 2009 after publication of this report. The findings of the survey will feed into forthcoming research and development work to be funded by WRAP that will assess the practicalities and economics of producing containers from mixed colour cullet.


Contents 1.0 Introduction ............................................................................................................................. 8

1.1 Mixed colour cullet.................................................................................................................8 1.1.1 Glass colour match slides ..........................................................................................9

1.2 Consumer perception samples................................................................................................9 1.2.1 Sample preparation...................................................................................................9

2.0 Consumer perception study ................................................................................................... 10 2.1 Study context......................................................................................................................10 2.2 Research questions..............................................................................................................10 2.3 Products .............................................................................................................................11 2.4 Study One - “as is” vs “dark”................................................................................................13

2.4.1 Participants ............................................................................................................13 2.4.2 Materials ................................................................................................................13 2.4.3 Procedure...............................................................................................................14 2.4.3.1 Pre-rating questionnaire and rating questionnaire .....................................................14 2.4.3.2 Focus group ...........................................................................................................14

2.5 Study One – results .............................................................................................................15 2.5.1 Effects of glass colour on ratings for all products ......................................................15 2.5.2 Effects of glass colour on ratings by product.............................................................15 2.5.3 Significant colour by age interactions .......................................................................20 2.5.4 Significant colour by gender interactions ..................................................................23 2.5.5 Focus group results Study One ................................................................................28 2.5.5.1 Pinot Bianco ...........................................................................................................28 2.5.5.2 Chardonnay............................................................................................................28 2.5.5.3 Basics jam..............................................................................................................28 2.5.5.4 Regular jam............................................................................................................28 2.5.5.5 Whisky ...................................................................................................................29 2.5.5.6 Regular mayonnaise................................................................................................29 2.5.5.7 Basics mayonnaise ..................................................................................................29 2.5.5.8 Pasta sauce ............................................................................................................29 2.5.5.9 Gin.........................................................................................................................29 2.5.5.10 Comments on recycling and marketing of ‘green glass’..............................................29

2.6 Study One – discussion ........................................................................................................30 2.7 Study Two – “as is” vs “light” ...............................................................................................31

2.7.1 Participants ............................................................................................................31 2.7.1.1 Study Two..............................................................................................................32

2.8 Study Two – results .............................................................................................................32 2.8.1 Effects of glass colour on ratings for all products ......................................................32 2.8.2 Effects of glass colour on ratings by product.............................................................33 2.8.3 Significant colour by age interactions .......................................................................37 2.8.4 Significant colour by gender interactions ..................................................................38 2.8.5 Study Two – focus group results ..............................................................................41 2.8.5.1 Pinot Bianco ...........................................................................................................41 2.8.5.2 Chardonnay............................................................................................................41 2.8.5.3 Rosé ......................................................................................................................41 2.8.5.4 Basics jam..............................................................................................................41 2.8.5.5 Regular jam............................................................................................................42 2.8.5.6 Whisky ...................................................................................................................42 2.8.5.7 Regular mayonnaise................................................................................................42 2.8.5.8 Basics mayonnaise ..................................................................................................42 2.8.5.9 Pasta sauce ............................................................................................................42 2.8.5.10 Gin.........................................................................................................................43 2.8.5.11 Comments on recycling and marketing of ‘green glass’..............................................43

2.9 Study Two – discussion........................................................................................................43 3.0 Conclusion.............................................................................................................................. 45 4.0 Demand from retailers and brand owners ............................................................................. 46


Appendix 1:Pre-rating questionnaire and rating questionnaire ........................................................ 47 Appendix 2: Products ......................................................................................................................... 52 Figures Figure 1 Examples of the colour match slides produced. ..................................................................................9 Figure 2 Schematic outline of the study including approximate time for each section.......................................13 Figure 3 (a) Gender of the participants in Study One and (b) Age of the participants in Study One. .................13 Figure 4 Example products used in the Study One.........................................................................................14 Figure 5 Mean ratings for all products in both glass colours. ..........................................................................15 Figure 6 Mean ratings for each item for Pinot Bianco in both glass colours......................................................16 Figure 7 Mean ratings for each item for Chardonnay in both glass colours. .....................................................16 Figure 8 Mean ratings for each item for Basics Jam in both glass colours........................................................16 Figure 9 Mean ratings for each item for Regular Jam in both glass colours. ....................................................17 Figure 10 Mean ratings for each item for Whisky in both glass colours. ..........................................................17 Figure 11 Mean ratings for each item for Regular Mayonnaise in both glass colours. .......................................17 Figure 12 Mean ratings for each item for Basics Mayonnaise in both glass colours. .........................................18 Figure 13 Mean ratings for each item for Pasta Sauce in both glass colours....................................................18 Figure 14 Mean ratings for each item for Gin in both glass colours.................................................................19 Figure 15 Colour by Age interaction for ‘High Quality’ for Basics Jam. ............................................................20 Figure 16 Colour by Age interaction for the rating item ‘I think the appearance of this bottle / jar is pleasing’ for Basics Jam. ..................................................................................................................................................20 Figure 17 Colour by Age interaction for the rating ‘I think the contents of this bottle / jar are fresh’ for Basics Jam. ............................................................................................................................................................21 Figure 18 Colour by Age interaction for the rating ‘I would like to purchase this product’ for Whisky................21 Figure 19 Colour by Age interaction for the rating ‘I think the contents of this bottle / jar will be strong tasting’ for Chardonnay.............................................................................................................................................21 Figure 20 Colour by Age interaction for the rating ‘I think this bottle / jar is typical for a product of this type’ for Gin. .............................................................................................................................................................22 Figure 21 Colour by Age interaction for the rating ‘I think the contents of this bottle / jar are Fresh’ for Gin.....22 Figure 22 Colour by Gender interaction for the rating ‘I think the contents of this bottle / jar are Appetising’ for Jam from the Basics range. ...........................................................................................................................23 Figure 23 Colour by Gender interaction for the rating Purchase Likelihood for Mayonnaise from the Regular range...........................................................................................................................................................23 Figure 24 Colour by Gender interaction for the rating ‘I think this product is High Quality’ for Mayonnaise from the Regular range.........................................................................................................................................23 Figure 25 Colour by Gender interaction for the rating ‘I think the contents of this bottle / jar are Pure’ for Mayonnaise from the Regular range...............................................................................................................24 Figure 26 Colour by Gender interaction for the rating ‘I think the appearance of this bottle / jar is Typical for this product’ for Mayonnaise from the Regular range.............................................................................................24 Figure 27 Colour by Gender interaction for the rating ‘I think the appearance of this bottle / jar is Pleasing’ for Mayonnaise from the Regular range...............................................................................................................24 Figure 28 Colour by Gender interaction for the rating ‘I think the contents of this bottle / jar are Appetising’ for Mayonnaise from the Regular range...............................................................................................................25 Figure 29 Colour by Gender interaction for the rating ‘I think the contents of this bottle / jar are Fresh’ for Mayonnaise from the Regular range...............................................................................................................25 Figure 30 Colour by Gender interaction for the rating ‘I think the contents of this bottle / jar will be Strong Tasting’ for Mayonnaise from the Regular range. ............................................................................................26 Figure 31 Colour by Gender interaction for the rating ‘I think the appearance of this bottle / jar is Typical for this product’ for Mayonnaise from the Basics range...............................................................................................26 Figure 32 Colour by Gender interaction for the rating ‘I think the appearance of this bottle / jar is Pleasing’ for Mayonnaise from the Basic range. .................................................................................................................27 Figure 33 Colour by Gender interaction for the rating ‘I think the appearance of this bottle / jar are Fresh for Mayonnaise from the Basic range. .................................................................................................................27 Figure 34 Colour by Gender interaction for the rating ‘It is easy for me to tell whether I would like the contents of this bottle / jar’ for Pasta sauce. ................................................................................................................28 Figure 35 (a) Gender of the participants in Study Two and (b) Age of the participants in Study Two................31 Figure 36 Example products used in Study Two. ...........................................................................................32 Figure 37 Mean ratings for all products in both glass colours. ........................................................................32 Figure 38 Mean ratings for each item for Pinot Bianco in both glass colours ...................................................33


Figure 39 Mean ratings for each item for Chardonnay in both glass colours. ...................................................33 Figure 40 Mean ratings for each item for Rosé wine in both glass colours.......................................................34 Figure 41 Mean ratings for each item for Basics Jam in both glass colours......................................................34 Figure 42 mean ratings for each item for Regular Jam in both glass colours. ..................................................34 Figure 43 Mean ratings for each item for Whisky in both glass colours. ..........................................................35 Figure 44 Mean ratings for each item for Regular Mayonnaise in both glass colours. .......................................35 Figure 45 Mean ratings for each item for Basics Mayonnaise in both glass colours. .........................................35 Figure 46 Mean ratings for each item for Pasta Sauce in both glass colours....................................................36 Figure 47 Mean ratings for each item for Gin in both glass colours.................................................................36 Figure 48 Colour by Age interaction for the rating ‘I think the contents of this bottle / jar are Fresh’ for regular Mayonnaise. .................................................................................................................................................37 Figure 49 Colour by Age interaction for the rating ‘It is easy for me to tell whether I would like the contents of this bottle /jar’ for regular Jam. .....................................................................................................................37 Figure 50 Colour by Age interaction for the rating ‘I think this product is attention grabbing’ for Mayonnaise from the Basics range. ..........................................................................................................................................38 Figure 51 Colour by Gender interaction for the rating ‘I think this product is Attention Grabbing’ for Pinot Bianco....................................................................................................................................................................38 Figure 52 Colour by Gender interaction for the rating ‘I think the appearance of this bottle / jar is Typical for a product of this type’ for Pinot Bianco. ............................................................................................................38 Figure 53 Colour by Gender interaction for the rating ‘I would like to purchase this product’ for Jam from the Basic range. .................................................................................................................................................39 Figure 54 Colour by Gender interaction for the rating ‘I think this product is High Quality’ for Pasta Sauce. ......39 Figure 55 Colour by Gender interaction for the rating ‘I think the contents of this bottle / jar are appetising’ for Pasta Sauce. ................................................................................................................................................39 Figure 56 Colour by Gender interaction for the rating ‘I think the contents of this bottle / jar are appetising’ for Jam from the Regular range..........................................................................................................................40 Figure 57 Colour by Gender interaction for the rating ‘I think the contents of this bottle / jar are Fresh’ for Jam from the regular range..................................................................................................................................40 Figure 58 Colour by Gender interaction for the rating ‘I think this product is High Quality’ for Mayonnaise from the Basic range.............................................................................................................................................40 Tables Table 1 Average colour composition of mixed colour cullet (2008) at the point of delivery to a glass processor. ..8 Table 2 Order of product presentation. .........................................................................................................12


Glossary Term Definition

Cullet Recycled glass.

External cullet Recycled glass collected from bottle banks and kerbside.

Internal cullet Reject glass generated during the manufacturing process that is returned to the furnace.

Likert scale A psychometric scale commonly used in questionnaires and survey research. Respondents specify their level of agreement to a statement.

Acknowledgements GTS would like to acknowledgement the support and assistance of the following of which this project would have not been successful:

Paula Chin, Sainsbury’s, for support and provision of samples for the study;

Tim Histed, Store Manager, Sainsbury’s White Rose Shopping Centre and his staff for the provision of a

testing area and supporting the researchers;

Ed Sutherland, Liz Edwards and Steve Westerman, University of Leeds for undertaking the consumer

perception study;

Tammy Marrett, WRAP, for support and project management;

Monika Laszczynska and Colin Waller, Bottleworks Ltd, for spray coating of the samples; and

GTS staff for preparing the samples.

Assessing the demand for glass containers with maximum recycled content

8

1.0 Introduction The objective of this project is to assess the marketability to consumers and the demand from retailers and brand owners for glass containers with a maximum glass recycled content made from mixed colour cullet. In recent years the proportion of cullet collected in a mixed colour format as oppose to colour separated has increased significantly. This is in part due to a notable increase in domestic kerbside and hospitality sector collections, the majority of which collect the glass in a mixed colour format. This trend has been accompanied by a decrease in traditional bottle bank collections which predominately collect glass as separate colours. Assuming that the growth in mixed colour glass collections will continue, then less cullet colour separated at source will be available for remelt back into new glass containers; remelt of cullet into new glass containers represents closed loop recycling and offers the best environmental option for the use of container cullet. In addition there is currently an imbalance between the colours of glass that are recycled from the UK waste stream, and the colour demanded by UK container manufacturers. In 2008 around half of the glass recycled in the UK was green with flint making up 37.5% and amber at around 11.5%. The high volume of green glass in the UK waste stream is attributed to the large volumes of imported green containers, particularly wine. UK container glass production is flint 64%, amber 20% and green at 16% and flint is predominantly for the food and spirit sectors. The result of this scenario is that the UK has become a net importer of green glass much of which is currently going into secondary markets such as aggregates. By challenging consumer expectation this work could stimulate a market within the container glass industry for this green cullet. Two approaches to address the increasing proportion of cullet being collected as mixed colour are greater use of existing colour separation technologies to produce single colour glass streams, and increased manufacture of glass containers from colour mixed rather than single colour cullet. This project investigates the latter option with regard to the likely impact on consumers buying decisions of manufacturing glass containers from mixed colour rather than colour separated cullet. More specifically, this project investigated differences in consumer perceptions for a range of food and drink products traditionally packed in flint (clear) glass, when packed in glass that was green due to incorporation of high levels of mixed colour cullet. 1.1 Mixed colour cullet There are two main sources of mixed colour cullet:

Municipal Sector: glass mainly collected from domestic households and some limited quantity from commercial

premises; and

Hospitality Sector: glass collected from premises such as pubs, clubs, restaurants and hotels.

The colour composition of these two cullet sources varies due to the different mix of glass containers consumed in these different sectors. Glass from domestic households tends to have higher levels of green glass due to home consumption of wine and beer and a legacy effect that bring sites were originally called ‘bottle banks’ and in consequence jars that tend to be of flint glass, remain less likely to be recycled. Conversely glass from the hospitality sector tends to be mainly a mix of flint and green, with a higher proportion of flint glass. In order to characterise the colour composition by sector, the UK’s main glass processors provided colour compositional data for a 12 month period. Table 1 shows the average colour composition and the range for 2008.

Table 1 Average colour composition of mixed colour cullet (2008) at the point of delivery to a glass processor. Colour Municipal Hospitality

Average Range Average Range

Flint 30% 22% to 35% 48% 40% to 58%

Green 59% 49% to 71% 41% 36% to 44%

Brown 11% 7% to 11% 11% 6% to 20%


It is noted that colour composition does also vary depending on the time of year and geographical region. In consequence, the colour of glass made from mixed colour cullet will also vary if no colour specification is applied. 1.1.1 Glass colour match slides In order to determine the glass colour resulting from the melting of mixed colour cullet, physical glass samples were prepared based on the average colour compositions in Table 1 for both municipal and hospitality sectors. For the propose of this study it was agreed that the recycled content would be 80%. Therefore, mixed cullet of the appropriate composition was mixed with 10% raw materials to produce approximately 1 kg of glass. This mix represents 80% external cullet with 10% internal cullet and 10% raw materials, and using WRAP’s current cullet recycled content definition1 gives a glass with 80% recycled content. The cullet and raw material mixture was melted at 1450°C and cast as 40mm disks. The disks were ground and polished to produce colour slide disks at 1, 2, 3 and 4mm thickness ( Figure 1). The glasses containing municipal and hospitality cullet were termed “dark” and hospitality as “light” respectively.

Figure 1 Examples of the colour match slides produced. 1.2 Consumer perception samples Sainsbury’s and WRAP agreed that the following products would be investigated in the consumer perception study:

Sainsbury’s strawberry jam;

Sainsbury’s Basics strawberry jam;

Sainsbury’s mayonnaise;

Sainsbury’s Basics mayonnaise;

Sainsbury’s tomato and herb pasta sauce;

Sainsbury’s Pinot Bianco white wine;

Sainsbury’s Chardonnay white wine;

Sainsbury’s London dry gin;

Sainsbury’s Scotch whisky; and

Sainsbury’s Spanish rosé.

1.2.1 Sample preparation It is in principle technically feasible to manufacture glass containers at an industrial scale with maximum mixed colour cullet content. However, until it is demonstrated that there is market demand for such containers it is not within the financial constraints of the project to manufacture such containers. To put this into context, if a typical 300 tonnes/day furnace producing green glass was used for the manufacture of trial containers, it would take at least two days to flush existing conventional green glass though and allow the glass colour to stabilise. A further

1 Glass Toolkit - Recycled content toolkit, http://www.wrap.org.uk/retail/materials/glass_recycled_content_toolkit/

“light” Hospitality

“dark” Municipal


day would be required to produce the trial containers, followed by three days to allow the furnace to return back to normal green production. This process would equate to the potential loss of a week’s production, and at 90% cullet addition there could be nearly 2,000 tonnes of cullet used that would be flushed out and would be unsuitable for subsequent container manufacture. A final consideration is that green furnaces are fitted with IS Machines (container making machines) that are set up for bottle production and therefore it would be not be possible to produce food jars. If food jars were required then a flint furnace with IS Machines setup for jar production would need to be run using mixed colour cullet. Returning a flint furnace back to normal would run the risk of high reject levels as flint glass is much more susceptible to colour contamination than green glass. In order to avoid the cost and commercial issues associated with the use of a full scale container production unit for the production of relatively small number of sample containers, the “light” and “dark” glass colours were simulated using transparent spray coatings applied by Bottleworks Ltd. Such coatings are used commercially to change the colour of glass bottles and to produce other finish effects, and in this case was used to simulate bottles and jars produced from mixed colour cullet. All the bottles and jars spray coated were flint glass, however, there was a slight variation in colour of the virgin flint glass due to the bottles and jars being produced by different glass manufacturers. In consequence this slightly affected the final shade of the sprayed bottles and jars. For each of the products considered in this study thirty samples were taken, and their packaging treated as follows:

10 samples were kept in their original packaging (no coating);

10 were given a “light” colour coating to represent bottles and jars containing hospitality cullet; and

10 were given a “dark” colour coating to represent bottles and jars containing municipal cullet.

The coated bottles and jars had their contents and label removed prior to spray coating and were then refilled and relabelled. 2.0 Consumer perception study 2.1 Study context The inclusion of higher levels of recycled glass results in glass that is greener in colour than traditional flint glass and the aim of this research was to investigate the effect of this greener glass on consumers’ perceptions on a range of Sainsbury’s products presented in bottles and jars with high mixed colour cullet recycled content. The “light” and “dark” shades were presented to consumers in two separate studies. The aim of this study is to provide information to retailers, brand owners and glass manufacturers in order for informed decisions about use of glass containers with high mixed colour glass recycled content in the retail sector. 2.2 Research questions Three main questions to be addressed by the research conducted were:

1) Does the “dark” shade of green glass influence the decisions that consumers make about purchase likelihood, quality and other perceptions of a product compared to the same product presented in flint glass?

2) Does the “light” shade of green glass influence the decisions that consumers make about purchase

likelihood, quality and other perceptions of a product compared to the same product presented in flint glass?

3) The research will also examine any effects of both age and gender of consumers and glass colour on the

perception of the products and how these factors interact with glass colour. These research questions were addressed through the use of both quantitative and qualitative approaches, namely a rating questionnaire and focus group discussions in which participants could raise issues of concern or interest and consider reasons for /influences on purchase decisions. Of particular interest here was the issue of how the reason for using green glass could best be communicated to consumers and whether this would be likely to influence their purchase decisions. This multi-method approach allows for triangulation of findings and


consequently greater confidence in the results. It also allowed evaluation of products in different contexts. For the ratings component of the studies participants were not made aware of the change in glass colour and products were presented in a simulated ‘on shelf’ scenario. However, participants were informed of the colour change prior to the focus group and products were presented in both ‘on the shelf’ and ‘in use’ (with some of the product removed) scenarios. Two separate studies were conducted. In Study One flint glass (“as is”) was compared to a darker green (“dark”) recycled glass for nine different products (18 containers in total, nine dark green and nine flint) with participants rating the viewed products and then taking part in a focus group to discuss the products. In Study Two flint glass (“as is”) was compared to a lighter green (“light”) recycled glass for 10 products (an additional product was added for Study Two making a total of 20 containers viewed, 10 light green and 10 flint). As in Study Two the products were rated and this was followed by a focus group. The two studies were constructed in this way as it allows direct comparison of flint glass to each of the colours of green glass without the need to show an unfeasibly large number of products to participants. 2.3 Products In Study One there were nine products in each of the two glass colours (“as is” and “dark”), with an additional product added for Study Two examining the “light” glass. The products were items sold by Sainsbury’s and were labelled as are the supermarket sold versions. Section 1.2 lists the products investigated in this study, the Sainsbury’s Spanish Rosé was only investigated in Study Two. Products were presented in two different orders with either green glass first or flint glass first, with the same order being used for the two glass colours, making four presentation orders in all (


Figure 2). This procedure ensured that the same product was never viewed consecutively in the two glass colours. Consequently order of presentation, flint glass first and green glass second or vice versa was a between participants variable. Order of presentation is shown below in Table 2. Products were presented in the order 1 -10 or 10-1 for different groups of participants.

Table 2 Order of product presentation.

No. Product 1 Pinot Bianco 2 Jam – Basics 3 Whiskey 4 Mayonnaise – Regular 5 Chardonnay 6 Pasta sauce 7 Gin 8 Jam – Regular 9 Rose wine (Study Two only) 10 Mayonnaise – Basics

For each study participants completed the following three elements:

1) a pre-rating questionnaire that collected demographic information;

2) a rating task in which all the products were presented, in a simulated ‘on shelf’ scenario, in the two

relevant glass colours and rated on a range of measures (see Appendix 1). Each item was rated on 10

different scales and participants were also required to estimate the price of each item. Finally

participants were asked what they believed the study was investigating. Presentation sequences were

counterbalanced to avoid order effects; and

3) a focus group during which participants’ views on products/containers were solicited. For these the items

were presented again, including ‘in use’ samples which had some of the product removed.


Male 35%

Female65%

18-3438%

35-5530%

56+32%

Figure 2 Schematic outline of the study including approximate time for each section. 2.4 Study One - “as is” vs “dark” 2.4.1 Participants A total of seventy-nine participants took part in Study One. Participants were recruited in a local Sainsbury’s supermarket. Participants were selected to represent the range of Sainsbury’s adult customers and consequently the sample consists of both males and females in three age ranges (18-34, 35-55, and 55+ years). A breakdown of participants’ gender and age is shown in Figure 3a (gender) and Figure 3b (age) respectively. Participants were paid £15 for their participation, which lasted approximately one hour and fifteen minutes.

Figure 3 (a) Gender of the participants in Study One and (b) Age of the participants in Study One. 2.4.2 Materials Study One involved presentation of 2x9 products (two glass colours and nine products). Products were Sainsbury’s branded versions of jam, mayonnaise, wine, whisky, gin and pasta sauce and included examples of both ‘standard’ ranges and ‘basic’ ranges for the jam and mayonnaise. Products were presented to participants in purpose built shelving units to simulate ‘on the shelf’ presentation (Figure 4). These shelving units restricted the

Study outline, ethical information and pre-rating questionnaire

Flint to Green Product Order 1 and product

ratings

Flint to Green Product Order 2 and product

ratings

Green to Flint Product Order 1 and product

ratings

Green to Flint Product Order 2 and product

ratings

Focus Group

10 mins

30 mins

30 mins

Total time approximately 1 hr 15 mins

5 mins Study Debrief

(a) (b)


amount of light from entering from the back and sides of the products as would be the case when the items were presented on the supermarket shelf. A questionnaire was used to measure the views of participants on 10 relevant ratings and an estimated cost for the item (see Appendix 1 for a copy of the questionnaire). These ratings included, but were not limited to purchase likelihood, whether the product appeared to be ‘fresh’, ‘appetising’, and ‘attention grabbing’. These ratings were measured on a seven point Likert scale with one representing ‘definitely not’ and seven representing ‘definitely’. Finally, in order to record the focus group discussions there was a video recorder and boundary microphone with supplementary notes made by the second experimenter during the discussion.

Figure 4 Example products used in the Study One. 2.4.3 Procedure 2.4.3.1 Pre-rating questionnaire and rating questionnaire On arrival in the testing room participants were told that the study involved rating a number of products and taking part in a focus group which would be recorded. Participants were told about ethical issues relating to the study and were required to sign an ethics form giving their informed consent. Following this, participants were required to complete the pre-rating questionnaire which collected demographic information relating to age, gender, household income, occupation and general shopping habits. Participants read instructions on how to complete the rating task, and these were reiterated by the experimenter. When all participants were ready to proceed they were presented with the different products, in one of the four presentation orders outlined in section 2.3. Participants then rated the 18 products on the 10 item rating scales and gave cost estimates. Each product was presented to the participants one at a time in a purpose built shelving unit with two products presented one behind the other so as to simulate on-shelf presentation as accurately as possible. Products were presented in two different orders with either green glass first or flint glass first, with the same order being used for the two glass colours, making four presentation orders in all. This procedure ensured that the same product was never viewed consecutively in the two glass colours. Consequently order of presentation, flint glass first and green glass second or vice versa was a between participants variable. Order of presentation is shown below in Table 2. Products were presented in the order 1-10 or 10-1 for different groups of participants. During the rating sessions participants were seated approximately one metre from the products and were allowed to take as much time as they required. 2.4.3.2 Focus group Following presentation of the rating scales a focus group was held. During these sessions the difference between the two glass colours, including the higher recycled content of the green glass, was explained to the participants and all the products were presented again, although some of them were presented in groups rather than individual products (wine, mayonnaise and jam). In the focus groups the green glass was presented alongside the flint glass so that a direct comparison between the two could be made. For all products participants were required to say whether they had originally noted the difference between the two glass colours, whether they could detect it when they saw the two products side by side, how they felt about the green glass and whether they would buy the product or not knowing that the glass had been recycled. Following presentation of all the products participants were asked about their general views on recycling and how the message about recycled glass could be effectively communicated to consumers.

Dark green Pinot Bianco Clear Pinot Bianco


0

1

2

3

4

5

6

Purchase At t ent ion Qualit y Pure Typical Pleasing Appet ising Fresh St rong Like

Flint Green

2.5 Study One – results 2.5.1 Effects of glass colour on ratings for all products This section outlines the main findings of the rating task element of Study One. The section reports findings on each of the items on the rating scales and how they are influenced by the colour of the glass in for each product presented. When a result is said to be ‘statistically significant’ at the 5% (.05) level then this means that there is a better than 95% probability the results are not due to chance.

Figure 5 Mean ratings for all products in both glass colours. Figure 5 shows the mean ratings across all products for both glass colours. The majority of items from the questionnaire showed a statistically significant difference between the two glass colours, as follows:

‘purchase likelihood’ there was a higher mean for products in flint glass rather than green. Statistical analysis

shows that this difference was significant (F=38.10, p<.001);

‘I think this product is attention grabbing’ there was a higher mean rating for products in green glass rather

than flint. Statistical analysis shows that this difference was significant (F=8.93, p<.001);

‘I think this product looks high quality’ there was a higher mean rating for products in flint glass rather than

green. This difference was statistically significant (F=10.98, p<.01);

‘I think this product looks pure’ there was a higher mean rating for products in flint glass rather than green.

This difference was statistically significant (F=10.94, p<.01);

‘I think this product looks typical’ for this product’ there was a higher mean rating for products in flint glass

rather than green. This difference was statistically significant (F=23.89, p<.001);

‘I think this product looks pleasing’ there was a higher mean rating for products in flint glass rather than

green. This difference was statistically significant (F=5.56 p<.05);

‘I think this product looks appetising’ there was a higher mean rating for products in flint glass rather than


‘I think this product looks fresh’ there was a higher mean rating for products in flint glass rather than green.


‘I think this product looks strong tasting’ there was a higher mean rating for products in green glass rather

than flint. However, this difference was not statistically significant (F=3.03, p=.086); and

‘it is easy for me to tell whether I would like the contents of this bottle / jar’ there was a higher mean rating

for products in flint glass rather than green. However, this difference was not statistically significant (F=1.41,

p=.238).

2.5.2 Effects of glass colour on ratings by product Previous section 2.5.1 Error! Reference source not found.reported the results for all products and showed that there were effects of glass colour on consumers’ perceptions of those products. However, these effects may be specific to certain products. Consequently this section reports the ratings for each of the individual products.


Figure 6 Mean ratings for each item for Pinot Bianco in both glass colours.

0

1

2

3

4

5

6

Me

an R

atin

g

Purchase Attention Quality Pure Typical Pleasing Appetising Fresh Strong Like

Flint Green

Figure 6 shows the mean ratings for each of the 10 items on the rating scale for Pinot Bianco for both flint and green glass (“dark”). Statistical analyses revealed that there were no significant differences for any of these items between the two glass colours for this product.

Figure 7 Mean ratings for each item for Chardonnay in both glass colours.

0

1

2

3

4

5

6

Mea

n R

atin

g


Flint Green

Figure 7 shows the mean ratings for each of the 10 items on the rating scale for Chardonnay for both flint and green glass (“dark”). Statistical analyses revealed that there were no significant differences for any of these items between the two glass colours for this product.

Figure 8 Mean ratings for each item for Basics jam in both glass colours.

0

1

2

3

4

5

6

Mea

n R

atin

g


Flint Green

Figure 8 shows the mean ratings for each of the 10 items on the rating scale for Basics jam for both flint and green glass (“dark”). Statistical analyses revealed that there was a significant difference on the rating ‘I think the appearance of this bottle / jar is pleasing’ with jam in the green jar rated as more pleasing than in the flint jar (F=4.188, p<.05). There were some differences that approached significance, namely the rating of purchase likelihood between the two glass colours (F=3.776, p=.056) with participants tending to be more likely to purchase the product in green glass rather than flint and the rating ‘I think the contents of this bottle / jar are fresh’ with jam in the green jar tending to be rated as appearing fresher than in the flint glass (F=3.812, p=.055). All other comparisons were non-significant.


Figure 9 Mean ratings for each item for Regular jam in both glass colours.

0

1

2

3

4

5

6

Mea

n R

atin

g


Flint Green

Figure 9 shows the mean ratings for each of the 10 items on the rating scale for Regular jam for both flint and green glass (“dark”). Statistical analyses revealed that there were no significant differences for any of these items between the two glass colours for this product. Figure 10 Mean ratings for each item for whisky in both glass colours.

0

1

2

3

4

5

6

7

Mea

n R

atin

g


Flint Green

Figure 10 shows the mean ratings for each of the 10 items on the rating scale for whisky for both flint and green glass (“dark”). Statistical analyses revealed that there were no significant differences for any of these items between the two glass colours for this product. Figure 11 Mean ratings for each item for Regular mayonnaise in both glass colours.

0

1

2

3

4

5

6

Mea

n R

atin

g


Flint Green

Figure 11 shows the mean ratings for each of the 10 items on the rating scale for Regular mayonnaise for both flint and green glass (“dark”). Statistical analyses revealed significant differences on most of the rating scales. Purchase likelihood was rated higher for mayonnaise in the flint jar than in the green jar (F=91.467, p<.001). Mayonnaise in the green jar was rated as more attention grabbing than in the flint jar (F=6.046, p<.05). Mayonnaise in the flint jar was perceived as being of higher quality than in the green jar (F=35.158, p<.001). Mayonnaise in the flint jar was considered to be purer than mayonnaise in the green jar (F=37.155, p<.001). The flint jar was considered to be more typical than the green jar for mayonnaise (F=37.412, p<.001). Mayonnaise in


the flint jar was considered to be more pleasing than in the green jar (F=28.276, p<.001). Mayonnaise in the flint jar was considered to be more appetising than in the green jar (F=89.204, p<.001). Finally, mayonnaise in the flint jar was considered to be fresher than it was in the green jar (F=38.515, p<.001). Other comparisons were non-significant. Figure 12 Mean ratings for each item for Basics mayonnaise in both glass colours.

00.5

11.5

22.5

33.5

44.5

5

Mea

n R

ati

ng


Flint Green

Figure 12 shows the mean ratings for each of the 10 items on the rating scale for Basics mayonnaise for both flint and green glass (“dark”). Statistical analyses revealed significant differences on most of the rating scales. Purchase likelihood was rated higher for flint rather than green glass (F=51.52, p<.001). The green mayonnaise jar was seen as more attention grabbing than the flint jar (F=16.36, p<.001). Mayonnaise in the flint jar was perceived as being of higher quality than in the green jar (F=14.22, p<.001). Mayonnaise in the flint jar was considered to be purer than mayonnaise in the green jar (F=26.32, p<.001). The flint glass jar was considered to be more typical than the green jar for mayonnaise (F=26.28, p<.001). Mayonnaise in the flint jar was considered to be more pleasing than in the green jar (F=5.39, p<.05). Mayonnaise in the flint jar was considered to be more appetising than in the green jar (F=56.759, p<.001). Mayonnaise in the flint jar was considered to be fresher than it was in the green jar (F=32.01, p<.001). Finally, Mayonnaise in the green jar was perceived as being stronger tasting than in the flint jar (F=6.09, p<.05). Figure 13 Mean ratings for each item for pasta sauce in both glass colours.

0

1

2

3

4

5

6

Mea

n R

atin

g


Flint Green

Figure 13 shows the mean ratings for each of the 10 items on the rating scale for pasta sauce for both flint and green glass (“dark”). Statistical analyses revealed that there was a significant difference between the two glass colours for the rating of purchase likelihood with pasta sauce in the flint jar more likely to be purchased than in the green jar (F=4.88, p<.05) and for the appetising rating with pasta sauce in the flint jar rated as more appetising than in the green jar (F=6.915, p<.05). All other comparisons were non-significant.


Figure 14 Mean ratings for each item for gin in both glass colours.

0

1

2

3

4

5

6

Mea

n R

atin

g


Flint Green


Figure 14 shows the mean ratings for each of the 10 items on the rating scale for gin for both flint and green glass (“dark”). Statistical analyses revealed that there were no significant differences for any of these items between the two glass colours for this product. 2.5.3 Significant colour by age interactions This section reports the statistically significant colour by age interaction effects for individual products. Figure 15 Colour by Age interaction for ‘high quality’ for Basics jam.

2

2.5

3

3.5

4

Flint Green

18-34

35-55

Over 55

Figure 16 Colour by age interaction for the rating item ‘I think the appearance of this bottle / jar is pleasing’ for Basics jam.

2

2.5

3

3.5

4

Flint Green

18-34

35-55

Over 55

Figure 15 shows the colour by age interaction for the rating ‘I think this product is high quality’ for jam from the Basics range. The graph shows no difference on this rating between the two glass colours for the 18-34 year olds while 35-55 year olds rated jam in the green jar as being of slightly lower quality than in the flint jar. However, the over 55 year olds rated jam in the green jar as being of higher quality than in the flint jar. Figure 16 shows the colour by age interaction for the rating ‘I think the appearance of this bottle / jar is pleasing’ for jam from the Basics range. The graph shows that 18-34 year olds rated jam in the green jar as more pleasing than in the flint jar, as did the over 55 year old group although the latter rated jam in both jar types as more pleasing than the 18-34 year olds did. However, the 35-55 year old group rated jam in the flint jar as more pleasing than in the green jar.


Figure 17 Colour by Age interaction for the rating ‘I think the contents of this bottle / jar are fresh’ for Basics jam.

2

2.5

3

3.5

4

4.5

5

Flint Green

18-34

35-55

Over 55

Figure 18 Colour by age interaction for the rating ‘I would like to purchase this product’ for whisky.

2

2.5

3

3.5

4

4.5

5

Flint Green

18-34

35-55

Over 55

Figure 17 shows the colour by age interaction for the rating ‘I think the contents of this bottle / jar are fresh’ for jam from the Basics range. The graph shows that the over 55 year old group rated jam in the green jar as appearing fresher than in the flint jar, as did the 18-34 year olds, although the difference between the two glass colours was smaller for the latter of the two groups. The 35-55 year old groups rated jam in the flint jar as appearing to be fresher than in the green jar, although again this difference was small. Figure 18 shows the colour by age interaction for the rating ‘I would like to purchase this product’ for whisky. The graph shows little difference on this rating between the two glass colours for both the 18-34 year olds and the 35-55 year olds. However, the over 55 year olds rated the purchase likelihood of the whisky as being higher when it was in the flint bottle. Figure 19 Colour by Age interaction for the rating ‘I think the contents of this bottle / jar will be strong tasting’ for Chardonnay.

4

4.5

5

5.5

6

Flint Green

18-34

35-55

Over 55


Figure 20 Colour by age interaction for the rating ‘I think this bottle / jar is typical for a product of this type’ for gin.

4

4.5

5

5.5

6

Flint Green

18-34

35-55

Over 55

Figure 19 shows the colour by age interaction for the rating ‘I think the contents of this bottle / jar will be strong tasting’ for Chardonnay. The graph shows that both the 18-34 year olds and the over 55 year olds rated Chardonnay in the green bottle as looking stronger than in the flint bottle, although the difference between the two bottles was smaller for the over 55 group. However, the 35-55 year olds rated Chardonnay as appearing stronger in the flint bottle. Figure 20 shows the colour by age interaction for the rating ‘I think this bottle / jar is typical for a product of this type’ for gin. The graph shows that the 18-34 year olds rated the green bottle as more typical than flint while the reverse was true for the 35-55 year old group. There was little difference on this rating between the two glass colours for the over 55 year olds. Figure 21 Colour by Age interaction for the rating ‘I think the contents of this bottle / jar are fresh’ for gin.

3

3.5

4

4.5

5

5.5

6

Flint Green

18-34

35-55

Over 55

Figure 21 shows the colour by age interaction for the rating ‘I think the contents of this bottle / jar are fresh’ for gin. The graph shows that the 18-24 year olds and the over 55 year old group both rated gin in the green bottle as fresher than in the flint bottle, whereas the 35-55 year old group rated gin in the flint bottle as fresher than in the green bottle.


2.5.4 Significant colour by gender interactions Figure 22 Colour by gender interaction for the rating ‘I think the contents of this bottle / jar are appetising’ for jam from the Basics range.

2

2.5

3

3.5

4

Flint Green

Male

Female

Figure 23 Colour by gender interaction for the rating purchase likelihood for mayonnaise from the Regular range.

11.5

22.5

33.5

44.5

5

Flint Green

Male

Female

Figure 22 shows the colour by gender interaction for the rating ‘I think the contents of this bottle / jar are appetising’ for jam from the Basics range. The graph shows that males rated jam in the green jar as more appetising than in the flint jar; there was little difference females. Figure 23 shows the colour by gender interaction for the rating of purchase likelihood for mayonnaise from the Regular range. The graph shows that both males and females rated the purchase likelihood of mayonnaise in the green jar as lower than they did for mayonnaise in the flint jar and that the effect was stronger in females than males. Figure 24 Colour by gender interaction for the rating ‘I think this product is high quality’ for mayonnaise from the Regular range.

11.5

22.5

33.5

44.5

5

Flint Green

Male

Female


Figure 25 Colour by gender interaction for the rating ‘I think the contents of this bottle / jar are pure’ for mayonnaise from the Regular range.

11.5

22.5

33.5

44.5

5

Flint Green

Male

Female

Figure 24 shows the colour by gender interaction for the rating ‘I think this product is high quality’ for mayonnaise from the Regular range. The graph shows that both males and females rated the mayonnaise in the green jar as lower quality than mayonnaise in the flint jar. Again, this effect was greater for females than males. Figure 25 shows the colour by gender interaction for the rating ‘I think the contents of this bottle / jar are pure’ for mayonnaise from the Regular range. The graph shows that both males and females rated mayonnaise in the green jar as less pure than in the flint jar. Again, this effect was greater for females than for males. Figure 26 Colour by gender interaction for the rating ‘I think the appearance of this bottle / jar is Typical for this product’ for mayonnaise from the Regular range.

22.5

33.5

44.5

55.5

6

Flint Green

Male

Female

Figure 27 Colour by gender interaction for the rating ‘I think the appearance of this bottle / jar is pleasing’ for mayonnaise from the Regular range.

11.5

22.5

33.5

44.5

5

Flint Green

Male

Female


Figure 26 shows the colour by gender interaction for the rating ‘I think the appearance of this bottle / jar is typical for this product’ for mayonnaise from the Regular range. The graph shows that both males and females rated the green jar as less typical than the flint jar, and that this effect was greater for females than males. Figure 27 shows the colour by gender interaction for the rating ‘I think the appearance of this bottle / jar is pleasing’ for mayonnaise from the regular range. The graph shows that both males and females rated the green jar as less pleasing than the flint jar and again this effect was greater for females than males. Figure 28 Colour by gender interaction for the rating ‘I think the contents of this bottle / jar are appetising’ for mayonnaise from the Regular range.

11.5

22.5

33.5

44.5

5

Flint Green

Male

Female

Figure 29 Colour by gender interaction for the rating ‘I think the contents of this bottle / jar are fresh’ for mayonnaise from the Regular range.

2

2.5

3

3.5

4

4.5

5

Flint Green

Male

Female

Figure 28 shows the colour by gender interaction for the rating ‘I think the contents of this bottle / jar are appetising’ for mayonnaise from the Regular range. The graph shows that both males and females rated mayonnaise in the green jar as less appetising than in the flint jar. The effect was larger in females than males. Figure 29 shows the colour by gender interaction for the rating ‘I think the contents of this bottle / jar are fresh’ for mayonnaise from the Regular range. The graph shows that both males and females rated mayonnaise in the green jar as less fresh than in the flint jar. The effect was larger in females than males.


Figure 30 Colour by gender interaction for the rating ‘I think the contents of this bottle / jar will be strong tasting’ for mayonnaise from the Regular range.

2

2.5

3

3.5

4

4.5

5

Flint Green

Male

Female

Figure 31 Colour by gender interaction for the rating ‘I think the appearance of this bottle / jar is typical for this product’ for mayonnaise from the Basics range.

2

2.5

3

3.5

4

4.5

5

5.5

Flint Green

Male

Female


Figure 30 shows the colour by gender interaction for the rating ‘I think the contents of this bottle / jar will be strong tasting’ for mayonnaise from the regular range. The graph shows that males rated the mayonnaise in the green jar as appearing stronger tasting than in the flint jar. However, females rated mayonnaise in the flint jar as appearing to have a stronger taste than in the green jar. Figure 31 shows the colour by gender interaction for the rating ‘I think the appearance of this bottle / jar is typical for this product’ for mayonnaise from the Basics range. The graph shows that both males and females rated mayonnaise in the green jar as less typical than in the flint jar and that this effect was larger in females than males. Figure 32 Colour by Gender interaction for the rating ‘I think the appearance of this bottle / jar is Pleasing’ for Mayonnaise from the Basic range.

1

1.5

2

2.5

3

3.5

4

Flint Green

Male

Female

Figure 33 Colour by gender interaction for the rating ‘I think the appearance of this bottle / jar are fresh for mayonnaise from the Basic range.

2

2.5

3

3.5

4

4.5

5

Flint Green

Male

Female

Figure 32 shows the colour by gender interaction for the rating ‘I think the appearance of this bottle / jar is pleasing’ for mayonnaise from the Basic range. The graph shows that males show little difference on this rating between the two glass colours whereas females rated mayonnaise in the green jar as less pleasing than in the flint jar. Figure 33 shows the colour by gender interaction for the rating ‘I think the contents of this bottle / jar are fresh’ for mayonnaise from the Basic range. The graph shows that both males and females rated mayonnaise in the green jar as less fresh than in the flint jar. The effect was larger in females than males.


Figure 34 Colour by gender interaction for the rating ‘It is easy for me to tell whether I would like the contents of this bottle / jar’ for pasta sauce.

3

3.5

4

4.5

5

5.5

6

Flint Green

Male

Female

Figure 34 shows the colour by gender interaction for the rating ‘It is easy for me to tell whether I would like the contents of this bottle / jar’ for Pasta sauce. The graph shows that males rated the green jar as easier to evaluate the product in than the flint jar whereas the opposite was true for females. 2.5.5 Focus group results Study One This section includes comments from the focus group with the statistical evidence from Study One to support the comments. 2.5.5.1 Pinot Bianco Many of the participants did not notice the colour difference between the two glass colours and some expressed a preference for the green glass. Comments included ‘green glass makes the wine look more full-bodied’, ‘green glass makes the wine look higher quality / more alcoholic’, ‘green bottles are not uncommon for wine’ ‘green bottle looks nice / attractive’ ‘difference between green and flint is less noticeable in Pinot Bianco than it is in Chardonnay’. However, one participant commented that the reverse of this was true suggesting that the green glass was more obvious for the Pinot Bianco than for the Chardonnay. Overall, the comments can be summarised as reflecting the statistical outcome of there being no clear preference for one glass colour over another and there were numerous comments that were favourable towards green glass for this product. 2.5.5.2 Chardonnay Focus group comments on Chardonnay were generally mixed and included ‘green glass was more noticeable for chardonnay than pinot bianco although it would not influence my purchase decision’, ‘flint glass looks lighter / fresher’, ‘green glass looks higher quality’, ‘green glass makes the wine look more expensive’, ‘flint glass makes the wine look more appealing and more yellow’. Overall then these comments reflect a range of views with their being no clear preference for one glass colour over another. 2.5.5.3 Basics jam Comments from the focus group included some positive comments about the green glass jar such as ‘jam in the green jar looks richer / tastier’ being made by a number of participants, as well as the ‘green jar makes the jam look more natural’ and others saying that ‘jam should be dark’. However, not all the comments were positive with some participants noting that ‘the green jar makes the jam look mouldy’ and interestingly ‘the green jar could be used to mislead by making the jam look darker than it actually is’ and that the ‘flint jar makes it easier to see the product’. Ambivalent comments included ‘green jar makes the jam look too dark, but I might not notice when it was on the shelf’ and ‘I can hardly the notice the difference’. Finally, an interesting comment was that the difference between the two glass colours was more obvious for jam from the basics range than jam from the regular range. 2.5.5.4 Regular jam Again the comments from the focus groups suggested no clear preference for one glass over the other. Comments here included ‘the green glass makes the jam look like it has a richer, deeper flavour’, ‘flint glass makes the jam look fresher’, ‘pips are harder to see in the green glass, but this would not put me off’, ‘if the jar was more full then it would not be possible to see that the glass was green’. Overall these comments suggest that there is no consistent preference for green or flint glass.


2.5.5.5 Whisky Focus group comments here were mixed with a number of negative comments made about whisky in green glass, such as ‘whisky in the clear bottle looks more golden, richer, purer’, ‘green looks less golden’, ‘green doesn’t look quite right, would choose clear first’, ‘clear glass bottle looks fresher, better quality / purer’, ‘green glass makes the whisky look like a specimen / sample’. However, there were some positive and neutral comments about the green glass including ‘green glass makes the whisky look richer’, ‘green glass makes the whisky look darker, but it would not influence my purchase decision’ and ‘difference is hardly noticeable, and would not influence my purchase’. Overall then comments were not consistent and reflect the outcome of the rating scales. 2.5.5.6 Regular mayonnaise Comments here were broadly negative with several participants suggesting that mayonnaise in the green jar looked like it had ‘gone off’, additional similar comments included ‘green is associated with food that has gone off’, ‘it looks not pure’, ‘it looks bilious’, ‘it looks like floor polish’, ‘there will be a lot of customer resistance’, ‘green glass is misleading – I thought it was the product not the jar’. However, some comments suggested that participants may buy the product in green glass, including ‘I would buy it as a novelty’, ‘I would buy it if it was cheaper than the clear jar’, ‘would buy it if I was aware of why the glass was green’, ‘lots of marketing would be needed to make this acceptable. Overall then the comments reflect the statistics with lots of resistance evident towards mayonnaise in a green jar. 2.5.5.7 Basics mayonnaise As with the regular mayonnaise there were a range of negative comments about mayonnaise in a green jar, including ‘it looks vile / like slime / disgusting / off / awful’ ‘I would never buy it in a million years’, ‘looks like a poorer product in green jar’, ‘it looks unnatural / futuristic’, ‘mayonnaise has got to be white’, ‘I prefer the clear glass’. However, there were a limited number of comments or suggestions that might make the green jar acceptable. These included ‘I would buy it in the green glass’, ‘a picture of white mayonnaise on the jar might help’, ‘clear lid might make me buy it’, ‘I would need to taste it first’, ‘full plastic wrap might make it acceptable’. It was also noted by a number of participants that the white lid and label made the green more apparent on this product than the mayonnaise from the regular range. Overall, as might be expected the comments were negative and reflected the outcome of the rating scales. 2.5.5.8 Pasta sauce Focus group comments for this product were mixed with a number of participants making negative comments about this product in the green glass. These negative comments included ‘in the green jar it looks less appealing’, ‘in the clear jar it looks more appetising, richer, sweeter’, ‘the green jar makes it look ‘mouldy / nasty / stale’, ‘in green jar it looks like something else has been added’ ‘green glass makes it look brown and like it has fewer tomatoes’, green jar makes it look off. However, there were some positive comments about the product in the green jar and these included ‘green looks stronger / richer / tastier’, ‘would buy the green glass if I knew the reason why it was green’ ‘would but green if it was cheaper’. Overall, there were more negative comments about this product in the green glass than positive ones, reflecting the findings of the statistical tests. 2.5.5.9 Gin While the comments from the focus group reflected a range of opinions there were a number of comments that were favourable about this product when presented in a green bottle. These included ‘colour doesn’t matter, if anything I prefer the green’, ‘Gin is often in green bottles’, ‘green stands out more and looks more expensive and higher quality’, ‘green makes it look classier’, ‘clear glass looks cheap’. However, others preferred the flint glass and this was expressed in comments such as ‘green bottle makes it look like mouthwash’, ‘gin should be clear / pure’. The comments here again largely reflect the findings reported above in section 2.5.2. 2.5.5.10 Comments on recycling and marketing of ‘green glass’ The majority of participants expressed a positive attitude towards recycling and considered it important, although there were a small number of exceptions. These exceptions were largely due to doubts that local authorities were recycling all waste and that the process was complex and time consuming. With regard to the promotion and marketing of a potential glass change participants suggested that there would need to be a large amount of marketing to inform consumers of the changes. However, in accordance with the statistical findings and the product specific comments many considered that the alcohol products could go straight on to the shelf, but that the food products would need more marketing. Suggestions for how to achieve this included television adverts, aisle banners, on-shelf information and product labelling, all of which should push the ‘green credentials’ of the recycled glass. Some participants also suggested that there should be in-store tasting so that consumers were


aware that the products were the same as pre-change products and that there could be some price reduction for the ‘green products’ or initial promotion of them. However, a number of participants remained resistant to the mayonnaise products even if they knew that the reason for it was a generally positive one. 2.6 Study One – discussion This study was designed to investigate whether there were differences in consumers’ perceptions of a range of products when presented in two different glass colours, flint glass and green recycled glass (“dark” glass colour) to males and females across three age groups. When considering all products (taken together) there were statistically significant differences on the majority of questionnaire ratings that were used with a preference being indicated for the products when in flint glass containers. Only rating items ‘I think it would be easy to tell if I would like this product’ and ‘I think this product would be strong tasting’ showed no significant effects. However, this effect was clarified when ratings were considered separately for each product and when interactions between colour and age and colour and gender were examined. More detailed analysis of the data indicates that effects are strong for some products and weak or non-existent for others, and that age and sex of the consumer also exerted an influence on preference in some instances. The strongest differences were found for both the mayonnaise products (Basics and Regular). For mayonnaise from the regular range the flint jar was rated significantly higher on seven of the rating scales. Mayonnaise in the flint jar was rated as more likely to be purchased, as being of higher quality, as purer, as more typical a container for this product, as more pleasing, as more appetising and as being fresher than the mayonnaise in the green jar. However, mayonnaise in the green jar was rated as more attention grabbing than in the flint jar. Mayonnaise from the basics range showed the same pattern of results with the additional finding that mayonnaise in the green jar was considered to be a stronger tasting product than in the flint jar. The results for these two products can be interpreted as a negative effect of green glass on consumers’ perceptions. Also of interest here are the interaction effects which show that the effect of glass colour had a greater effect on the ratings made by females on a number of the ratings scales, suggesting that females were more sensitive to the colour change for this product. This may well have implications for a potential change of glass colour, particularly as 82% of females reported that they did the shopping compared to only 43% of males. Focus group comments were in line with these findings with common responses to the product when in the green glass container suggesting that it looked like it had ‘gone off’ or was spoiled in some way. These comments suggest that the colour of the container reflects on the product itself rather than just the container and it is clear that the green glass influences the decisions that consumers make about the products. Comments also suggested that the label and lid colour can influence perception of glass colour with a number of participants noting that the green colour was more pronounced on the basics range mayonnaise which had a white lid and label. For jam from the Basics range there was a difference between the two glass colours for ratings of ‘pleasing appearance’ such that the green jar was preferred. Effects of glass colour for ratings of purchase likelihood and ‘freshness’ also approached significance and pointed in the same direction. There were some comments from the focus groups that supported this, with participants noting that the jam looked richer or tastier in the green container. However, there were also some comments favouring the flint container, which (contrary to the ratings) was thought to make the product look fresher and also made the product easier to see. There were a number of significant interactions of colour by gender and colour by age relating to this product suggesting that these effects need to be qualified and providing possible explanations for contrary reports. Interactions between age group and glass colour for ‘high quality’, ‘pleasing appearance’, and ‘freshness’ suggest that the middle age group prefers the flint glass container for this product (i.e., any preference for the green container may be restricted to the younger and older age groups). An interaction between glass colour and gender also indicated that only males, on average, considered the product more appetising in green glass. These interactions suggest that there needs to be some consideration of the demographics of consumers for this product and what effect this will have on choices consumers make. When considering the pasta sauce a difference was found on two of the rating scale items, namely ‘I would like to purchase this product’ and ‘I think this product looks appetising’ with the flint jar rated as more likely to be purchased and as more appetising than the green jar. There was also a colour by gender interaction on the rating ‘It is easy for me to tell whether I would like this product’ with males rating the green jar as easier to evaluate than the flint jar and the opposite effect evident in females suggesting that gender is potentially an issue in relation to the glass colour used for this product. Again the statistical analyses were supported by focus group comments such as the pasta sauce looking like there had been something added or that it gone stale when presented in the green container. However, there were also some positive comments directed towards the green container with some participants thinking that it would be richer/stronger tasting.


Male 41%Female

59%

18-3436%

35-5532%

56+32%

For the alcohol products – Pinot Bianco, Chardonnay, gin, and whisky as well as the jam from the Regular Sainsbury’s range there was no statistically significant main effect of glass colour on any of the ratings scales. However, it should be borne in mind that traditional statistical method tests for presence rather than absence of a difference. The finding of no statistically significant difference in this sample is not equivalent to saying that there will be no difference in the population. Although it is hoped that results are indicative, they must be interpreted with caution. Particularly so, because for some of these products interactive effects and the comments from the focus groups indicate a polarisation of views, with some participants showing a preference for green and some for flint. This was particularly the case for the gin and the whisky. Consistent with focus group comments for the Basics jam, participants thought that the taste might be stronger or richer in the green glass, but that the contents might be fresher and would be easier to see in the flint glass. There were significant interactions between glass colour and age for gin for both ratings of typicality and freshness. Mean ratings for the 35-55 year old group were different to those for the other age groups in rating the flint bottle as more typical and having contents that appeared ‘fresher’. There was also an interaction of colour and age on purchase likelihood for Whisky with the glass colour making little difference to the two younger groups but the older group rating purchase likelihood higher for the flint bottle than the green bottle. Finally, there was an interaction between colour and age on the rating of whether a product would be strong tasting or not for Chardonnay with the 18-34 year olds rating the green bottle as containing a stronger tasting product as, to a lesser degree did the over 55 year old group while the 35-55 year old group rated the flint bottle as containing the stronger product. In summary, there was a strong negative reaction to mayonnaise (both Basics and Regular) in the green glass, and this was somewhat more pronounced for women. For jam there was a polarisation of views, with comments from the focus groups indicating that it was likely to be thought stronger or richer tasting in the green glass, but fresher and would be easier to see in the flint glass. There was a statistically significant advantage in ratings of ‘pleasing appearance’ for the Basics jam in the green container, and other effects that just failed to reach significance also pointed in this direction. Interactive effects (for Basics) suggested that a preference for green was more likely for men and did not (on average) include the middle age group (35-55 years). Comments from the focus group indicated that perception of glass colour was heightened by the white label of the basics range, and this might explain why effects on ratings seem to be stronger for this container. For pasta sauce, focus group comments were in keeping with those for the jam – the product looked stronger/richer tasting in green, but looked fresher and would be easier to see in the flint. Ratings indicated it was regarded as more appetising in the flint jar and was more likely to be purchased. There were no significant main effects of glass colour, but polarisation of views was also apparent for the whiskey and gin. Although the means were similar, this should not be taken to indicate that there were no strong opinions. The observed interactions suggest that ratings might be influenced by consumers’ age. Finally, there were no significant differences in ratings of the wine (although there was one interactive effect for chardonnay) and participants’ comments in the focus groups indicate that they were generally fairly accepting of these products in the green bottles. Given that these not an uncommon container for wine on supermarket shelves, familiarity may play some role in determining responses. 2.7 Study Two – “as is” vs “light” 2.7.1 Participants A total of 37 participants took part in Study Two, details of the gender and ages of the participants used in Study Two are shown in figures 34(a) and 34(b) respectively.

Figure 35 (a) Gender of the participants in Study Two and (b) Age of the participants in Study Two.

(a) (b)


2.7.1.1 Study Two The design of Study Two was similar to that for Study One. However, the comparison of glass colour was between flint glass (as used in Study One) and “light” green recycled glass (replacing the darker green glass from Study One) (Figure 36). An additional product was also included for evaluation. Rosé wine was added to the nine products from Study One, making a total of ten products, each of which was presented in both flint and green making 20 presentations in total (see Table 2).

Figure 36 Example products used in Study Two. 2.8 Study Two – results 2.8.1 Effects of glass colour on ratings for all products

Figure 37 Mean ratings for all products in both glass colours.

0

1

2

3

4

5

6


Flint Green

Figure 37 shows the mean ratings across all products for both glass colours. Many of the items from the questionnaire showed a statistically significant difference between the two glass colours across all of which showed a preference for flint glass containers over green glass containers, as follows:

‘purchase likelihood’ there was a higher mean rating for products in flint glass rather than green. Statistical

analysis shows that this difference was significant (F=23.822, p<.001);

‘I think this product is attention grabbing’ there was a higher mean rating for products in flint glass rather

than green. However, this difference was non-significant (F=.691, p=.412);

‘I think this product looks high quality’ there was a higher mean rating for products in flint glass rather than


‘I think this product looks pure’ there was a higher mean rating for products in flint glass rather than green.


‘I think this product looks typical for this product’ there was a higher mean rating for products in flint glass

rather than green. This difference was statistically significant (F=6.908, p<.05);

Light green Pinot Bianco Clear Pinot Bianco


‘I think this product looks pleasing the figure shows that there was a higher mean rating for products in flint

glass rather than green. This difference was statistically significant, F=4.309 p<.05.

‘I think this product looks appetising’ there was a higher mean rating for products in flint glass rather than


‘I think this product looks fresh’ there was a higher mean rating for products in flint glass rather than green.

However, this difference was non-significant (F=2.056 p=.162);

‘I think this product looks strong tasting’ there was a higher mean rating for products in flint glass rather than

green. However, this difference was non-significant (F=.581, p=.452); and

‘it is easy for me to tell whether I would like the contents of this bottle / jar’ there was a higher mean rating

for products in flint glass rather than green. However, this difference was non-significant (F=.968, p=.333).

2.8.2 Effects of glass colour on ratings by product Section 2.8.1 reported the results for all products and showed that there were effects of glass colour on consumers’ perceptions of those products. However, these effects may be specific to certain products. Consequently this section reports the ratings for each of the individual products.

Figure 38 Mean ratings for each item for Pinot Bianco in both glass colours

0

1

2

3

4

5

6

Mea

n R

atin

g


Flint Green

Figure 38 shows the mean ratings for each of the 10 items on the rating scale for Pinot Bianco for both flint and green glass (“light”). Statistical analyses revealed that there were no significant differences for any of these items between the two glass colours for this product.

Figure 39 Mean ratings for each item for Chardonnay in both glass colours.

0

1

2

3

4

5

6

Mea

n R

atin

g


Flint Green

Figure 39 shows the mean ratings for each of the 10 items on the rating scale for Chardonnay for both flint and green glass (“light”). Statistical analyses revealed that there were no significant differences for any of these items between the two glass colours for this product.


Figure 40 Mean ratings for each item for rosé wine in both glass colours.

0

1

2

3

4

5

6

Mea

n R

atin

g


Flint Green

Figure 40 shows the mean ratings for each of the 10 items on the rating scale for rosé wine for both flint and green glass (“light”). Statistical analyses revealed that there were no significant differences for any of these items between the two glass colours for this product.

Figure 41 Mean ratings for each item for Basics jam in both glass colours.

0

1

2

3

4

5

6

Me

an R

ati

ng


Flint Green

Figure 41 shows the mean ratings for each of the 10 items on the rating scale for Basics jam for both flint and green glass (“light”). Statistical analyses revealed that there were no significant differences for any of these items between the two glass colours for this product.

Figure 42 mean ratings for each item for Regular jam in both glass colours.

0

1

2

3

4

5

6

Mea

n R

ati

ng


Flint Green

Figure 42 shows the mean ratings for each of the 10 items on the rating scale for Regular jam for both flint and green glass (“light”). Statistical analyses revealed that there were no significant differences for any of these items between the two glass colours for this product.


Figure 43 Mean ratings for each item for whisky in both glass colours.

0

1

2

3

4

5

6

Mea

n R

atin

g


Flint Green

Figure 43 shows the mean ratings for each of the 10 items on the rating scale for whisky for both flint and green glass (“light”). Statistical analyses revealed that there were no significant differences for any of these items between the two glass colours for this product.

Figure 44 Mean ratings for each item for Regular mayonnaise in both glass colours.

0

1

2

3

4

5

6

Me

an R

atin

g


Flint Green

Figure 44 shows the mean ratings for each of the 10 items on the rating scale for Regular mayonnaise for both flint and green glass (“light”). Statistical analyses revealed significant differences on most of the rating scales. Purchase likelihood was rated higher for the flint jar rather than the green jar (F=29.456, p<.001). Mayonnaise in the flint jar was perceived as being of higher quality than mayonnaise in the green jar (F=26.773, p<.001). Mayonnaise in the flint jar was considered to be purer than mayonnaise in the green jar (F=29.920, p<.001). The flint jar was considered to be more typical than the green jar for mayonnaise (F=32.776, p<.001). Mayonnaise in the flint jar was considered to be more pleasing than in the green jar (F=23.131, p<.001). Mayonnaise in the flint jar was considered to be more appetising than in the green jar (F=33.406, p<.001). Finally, mayonnaise in the flint jar was considered to be fresher than it was in the green jar (F=24.635, p<.001). Other comparisons were non-significant.

Figure 45 Mean ratings for each item for Basics mayonnaise in both glass colours.

0

1

2

3

4

5

6

Mea

n R

ati

ng


Flint Green


Figure 45 shows the mean ratings for each of the 10 items on the rating scale for Basics mayonnaise for both flint and green glass (“light”). Statistical analyses revealed significant differences on most of the rating scales. Purchase likelihood was rated higher for the flint rather than the green jar (F=15.841, p<.001). Mayonnaise in the flint jar was perceived as being of higher quality than mayonnaise in the green jar (F=10.465, p<.01). Mayonnaise in the flint jar was considered to be purer than mayonnaise in the green jar (F=5.043, p<.05). The flint jar was considered to be more typical than the green jar for mayonnaise (F=15.740, p<.001). Mayonnaise in the flint jar was considered to be more pleasing than in the green jar (F=15.764, p<.001). Mayonnaise in the flint jar was considered to be more appetising than in the green jar (F=22.2, p<.001). Mayonnaise in the flint jar was considered to be fresher than it was in the green jar (F=8.273, p<.01). All other comparisons were non-significant.

Figure 46 Mean ratings for each item for pasta sauce in both glass colours.

0

1

2

3

4

5

6

Mea

n R

atin

g


Flint Green

Figure 46 shows the mean ratings for each of the 10 items on the rating scale for pasta sauce for both flint and green glass (“light”). Statistical analyses revealed that there was only one significant difference between the two glass colours for this product. The difference was found on the rating scale ‘It is easy to tell whether I would like the contents of this bottle/jar’ with the flint jar being rated higher than the green jar (F=9.08, p<.01). All other comparisons were non-significant.

Figure 47 Mean ratings for each item for gin in both glass colours.

0

1

2

3

4

5

6

Mea

n R

atin

g


Flint Green

Figure 47 shows the mean ratings for each of the 10 items on the rating scale for gin for both flint and green glass (“light”). Statistical analyses revealed that there was only one significant difference between the two glass colours for this product. The difference was found on the rating scale ‘I think the contents of this jar / bottle are Pure’ with Gin in the flint bottle rated as purer than in the green bottle (F=5.493, p<.05). All other comparisons were non-significant.


2.8.3 Significant colour by age interactions This section reports the statistically significant colour by age interaction effects for individual products.

Figure 48 Colour by age interaction for the rating ‘I think the contents of this bottle / jar are fresh’ for regular mayonnaise.

1

1.52

2.53

3.5

44.5

55.5

6

Flint Green

18-34

35-55

Over 55

Figure 49 Colour by age interaction for the rating ‘It is easy for me to tell whether I would like the contents of this bottle /jar’ for regular jam.

4

4.5

5

5.5

6

Flint Green

18-34

35-55

Over 55

Figure 48 shows the colour by age interaction for the rating ‘I think the contents of this bottle / jar are Fresh’ for regular mayonnaise. The graph shows that all three age groups rated mayonnaise in the green jar as less fresh than in the flint jar. The difference between the two glass colours was similar for the 35-55 year olds and the over 55 year olds, but the effect was greater for the 18-34 year olds with them rating mayonnaise in the green jar as far less fresh than in the flint jar. Figure 49 shows the colour by age interaction for the rating ‘It is easy for me to tell whether I would like the contents of this bottle /jar’ for regular jam. The graph shows that 35-55 year olds rated the green jar as making the contents easier to evaluate than the flint jar, while the over 55 year olds rated the flint jar as making it easier to evaluate the product. There was no difference between the two glass colours on this rating for the 18-34 year olds.


Figure 50 Colour by age interaction for the rating ‘I think this product is attention grabbing’ for mayonnaise from the Basics range.

1

1.52

2.53

3.5

44.5

55.5

6

Flint Green

18-34

35-55

Over 55

Figure 50 shows the colour by age interaction for the rating ‘I think this product is attention grabbing’ for mayonnaise from the Basics range. The graph shows that the 18-34 year old group and the 55 year old group both rated mayonnaise in the green jar as more attention grabbing than in the flint jar, though this effect was greater for the former group. Conversely, the 35-55 year old group rated mayonnaise in the green jar as less attention grabbing than the flint jar. 2.8.4 Significant colour by gender interactions

Figure 51 Colour by gender interaction for the rating ‘I think this product is attention grabbing’ for Pinot Bianco.

2

2.5

3

3.5

4

4.5

5

Flint Green

Male

Female

Figure 52 Colour by gender interaction for the rating ‘I think the appearance of this bottle / jar is typical for a product of this type’ for Pinot Bianco.

4

4.5

5

5.5

6

6.5

7

Flint Green

Male

Female

Figure 51 shows the colour by gender interaction for the rating ‘I think this product is attention grabbing’ for Pinot Bianco. The graph shows that males rated the wine in the green bottle as more attention grabbing than in the flint bottle, whereas females rated the green bottle as slightly less attention grabbing than the flint bottle.


Figure 52 shows the colour by gender interaction for the rating ‘I think the appearance of this bottle / jar is typical for a product of this type’ for Pinot Bianco. The graph shows that males rated the green bottle as more typical than the flint bottle whereas females rated the green bottle as less typical.

Figure 53 Colour by gender interaction for the rating ‘I would like to purchase this product’ for jam from the Basic range.

2

2.5

3

3.5

4

4.5

5

Flint Green

Male

Female

Figure 54 Colour by gender interaction for the rating ‘I think this product is high quality’ for pasta sauce.

3

3.5

4

4.5

5

Flint Green

Male

Female

Figure 53 shows the colour by gender interaction for the rating ‘I would like to purchase this product’ for jam from the Basic range. The graph shows that males rated purchase likelihood as higher for Jam in the flint jar than in the green jar. There was little effect for females. Figure 54 shows the colour by gender interaction for the rating ‘I think this product is high quality’ for pasta sauce. The graph shows that males rated the green jar as lower in quality than the flint jar whereas females rated the green bottle as slightly higher quality.

Figure 55 Colour by gender interaction for the rating ‘I think the contents of this bottle / jar are appetising’ for pasta sauce.

3

3.5

4

4.5

5

Flint Green

Male

Female


Figure 56 Colour by gender interaction for the rating ‘I think the contents of this bottle / jar are appetising’ for jam from the Regular range.

3

3.5

4

4.5

5

5.5

6

Flint Green

Male

Female

Figure 55 shows the colour by gender interaction for the rating ‘I think the contents of this bottle / jar are appetising’ for pasta sauce. The graph shows that males rated pasta sauce in the green jar as less appetising than in the flint jar. There was no effect for females. Figure 56 shows the colour by gender interaction for the rating ‘I think the contents of this bottle / jar are appetising’ for jam from the regular range. The graph shows that males rated jam in the green jar as less appetising than in the flint jar. There was little effect for females who rated jam in the green jar as slightly more appetising.

Figure 57 Colour by gender interaction for the rating ‘I think the contents of this bottle / jar are Fresh’ for jam from the Regular range.

3

3.5

4

4.5

5

5.5

6

Flint Green

Male

Female

Figure 58 Colour by gender interaction for the rating ‘I think this product is high quality’ for mayonnaise from the Basic range.

1

1.5

2

2.5

3

3.5

4

Flint Green

Male

Female

Figure 57 shows the colour by gender interaction for the rating ‘I think the contents of this bottle / jar is fresh for jam from the Regular range. The graph shows that males rated jam in the green jar as less fresh than in the flint jar. Females rated jam as slightly more fresh in the green jar than in the flint jar.


Figure 58 shows the colour by gender interaction for the rating ‘I think this product is high quality’ for mayonnaise from the Basic range. The graph shows that females rated mayonnaise in the green jar as less high quality than in the flint jar. There was little effect for males who rated jam in the green jar as slightly lower in quality. 2.8.5 Study Two – focus group results This section includes comments from the focus group with the statistical evidence from Study Two to support the comments. 2.8.5.1 Pinot Bianco Many of the participants did not notice the colour difference between the two glass colours. Three people could see the glass colour difference more in the Pinot than the Chardonnay although it was a ‘subtle’ difference. The difference in colour was ‘hardly/barely’ noticeable, both glass colours were ‘equally acceptable’ as it was possible to see ‘that the wine is clear’ in both glass colours. ‘Flint looks like it might be a lighter taste’ and slightly darker in the light green than the flint and therefore ‘might be a stronger flavour’. However, another participant said they ‘wouldn’t attribute the colour to the glass’. Generally people are ‘more accepting of coloured glass for liquids’ (as opposed to food products). 2.8.5.2 Chardonnay Focus group comments here were similar to those for the Pinot Bianco, although the glass colour difference may be ‘slightly more obvious’ with the Chardonnay. Some preferred the colour of the wine in the flint bottle because it ‘looks like it might be a lighter taste, more tempting’ as the green glass with the Chardonnay ‘conveys a more heavy, oaky flavour that is not as appealing‘. Another person thought the green bottle made the wine ‘look duller’ but another said that the flint bottle made the wine ‘look cheap’. Generally these comments were followed by assurance they would not affect purchase decisions and would ‘pick either one’ or ‘can’t see a problem with the green bottle’. One person stated they ‘happily try different products for drinks’ and another said that they were ‘more interested in alcoholic strength and country of origin for wines, and that generally ‘for alcoholic products (although in response to white wines) wouldn’t need to know about recycled green glass so much’. 2.8.5.3 Rosé Focus group comments revealed some strong opinions both for and against this particular product dependent on the colour of the bottle. For example, a number of participants said they couldn’t really tell the difference in glass colour and would be ‘equally happy with either’. Despite the green bottle making the wine look ‘darkish’ for a rosé, this likely wouldn’t affect purchase decisions, although if flint and green bottles were displayed together they ‘might be tempted to go for the flint’. Another said ‘once you’ve poured the product, it will look the same’. Some participants thought the darker one (i.e. the green) ‘looks better quality’, ‘more full-bodied’ and to be ‘more inclined to buy it as it looks like a red wine’, and the flint bottle looks ‘weaker or diluted’. In contrast someone else said they preferred the flint because ‘it conveys a lighter, fresh, summery taste’, another who stated the flint looked like it would have a ‘smoother taste’. Comments made may depend on whether the respondent is a rosé buyer/drinker. One rosé drinker preferred the flint, another non- rosé buyer/drinker preferred the green because it ‘looks darker/stronger’ and went on, ‘if I liked rosé I would buy the darker one’. The flint green comparison strongly conveys a difference in strength. A number of participants stated that ‘the green bottle makes it look like a cheap red wine, not like a rosé and it is off putting and I would look for a paler one’ and for one person this dark colour ‘was accentuated by having two bottles stacked one behind the other’. However, for rosé wine the brand and its’ country of origin apparently ‘makes a difference more than the glass colour’. ‘It looked (a wine) of questionable quality – just says Spanish, no grape variety, dark in colour’, and ‘everything about it says it will burn’. 2.8.5.4 Basics jam Comments from the focus groups suggested no clear preference for one glass colour over the other. ‘I wouldn’t notice the difference even when flint and green on shelf’ and ‘you get different shades in jam anyway’. It is ‘easier to tell the colour difference’ in the Basics range, ‘bit more noticeable’ as opposed to the Regular brand. Although no strong opinions, ‘you have to look hard to see the difference’ but flint slightly preferred. Where participants did notice the difference, ‘the light green glass does not do the jam any favours’. On the Basics jam the ‘white’ packaging, i.e. the labelling and lid, emphasize the green of the glass. Most of the participants said glass colour would not affect their purchase decisions.


2.8.5.5 Regular jam Both the Basics and the Regular jam were seen as ‘darker’ in the green glass which suggested the jam was ‘more fruity’, had ‘riper strawberries’ and a ‘stronger taste’, although the latter ‘might not necessarily be a good thing’ i.e. green glass may be less attention grabbing in this instance and appears to have affected purchase decisions for the Regular jam. The ‘green looks more like blackcurrant’. Although it was ‘harder to tell the glass colour difference in the Regular jam jars’ it did look ‘ a little darker’ whereas the flint jar ‘looked fresher’. More than one participant stated that the flint jars ‘allow you to see the contents better, i.e. you can see the seeds’, the flint looks ‘more expensive’ i.e. higher quality, in the Regular (but not the Basics) range. 2.8.5.6 Whisky Generally comments suggested that participants seemed to prefer the flint bottle or to have no preference between flint or green bottles. One person said ‘I probably prefer the flint because I am used to it’. Many preferred the flint bottle because it looked ‘brighter’, ‘clearer and purer’ like it was made with ‘clear spring water’, and having a ‘crisper, or sharper or cleaner’ taste. Some thought the whisky in the green bottle looked less appealing and one said it ‘looked like urine’. The green glass makes the whisky look ‘cheaper’ and appear to have a ‘harsher taste’. However, the green bottle was seen by some participants as looking ‘stronger’ and ‘more alcoholic’ and they preferred it to the flint. Indeed, when the flint and green bottles are not side by side, the green bottle ‘looks fine’. 2.8.5.7 Regular mayonnaise Comments from the focus group showed a clear preference for the mayonnaise in flint jars. The majority of participants disliked mayonnaise in the green jar with comments such as ‘green is associated with mould and I would be worried with eggs being an ingredient of mayonnaise’, and ‘I would never pick this up as it looks like it has gone off, and its not appetising’. Others thought that other ingredients might have been added to it including herbs or garlic. One person said ‘it’s not a nice green’, another said it ‘looks radioactive’. Other comments include it looks ‘toxic’, ‘a Halloween special’, and ‘like a paint pot’. A few participants said that knowing it was recycled glass ‘I’d go for the light green jar, but if I didn’t know I would go for the nicer packaged one (i.e. the Regular rather than the Basics) in flint because I would expect the taste to be better’. But one person (who had noticed the glass colour change in the ratings part of the study) said they liked it and thought it was ‘a richer product, perhaps because it was something different’. Overall, the focus group discussions indicated a negative view of this product in the light green glass. 2.8.5.8 Basics mayonnaise As with the Regular mayonnaise there was a clear preference for the mayonnaise in flint jars. The majority said that the green glass ‘put them off’ and ‘even if labelled as recycled glass I wouldn’t even pick it up!’ One person however thought the Basics jar ‘looked better in the light green glass’. Other participants thought people might get used to the light green jars quickly if it was made obvious what was causing the colour, saying ‘they would be able to see beyond the glass colour’, and ‘once you knew the product was the same I’d buy it’. However, purchase likelihood was affected by cost, with one person saying that ‘I’d probably buy in the green jar if I knew it was recycled but it depends on the price so that whichever was cheaper I would buy regardless of colour’. As with the Regular mayonnaise, the focus group discussions indicated there was resistance to this product in green glass although for this brand the price may have a stronger influence than glass colour on purchase decisions. Some participants also noted that the green was more obvious in the Basics range because of the white lid and label suggesting that the whole of the package and not just the glass colour needs to be considered. 2.8.5.9 Pasta sauce The overall tone of the comments from the focus groups suggested that the flint jar was preferred to the green jar for this product. Comments included: ‘flint looks more tomato-ey and more appetising’, and ‘the green one doesn’t look appetising’. The ‘darker’ colour was not well accepted by some participants as ‘it looked too brown’, ‘less tomato-ey’, looked ‘sludgy’ or ‘muddier’. Flint ‘looks fresher’ whereas the green jar ‘looks like it has been open for some time’ or ‘left in the sun’ or ‘it’s been on the shelf for a long time’. However, others thought the darker colour might be more concentrated and ‘have more tomatoes in it’ and another thought it might be ‘better quality, have better ingredients, and be more flavoursome’. There was also a significant difference on the rating scale ‘It is easy to tell whether I would like the contents of this bottle/jar’ with participants saying it was easier to evaluate the product in the flint jar such as ‘I like to see what I’m getting’. The ‘light packaging/label brings out the darkness of the product’ one said, and overall the flint jar ‘was better on the eye’. Further although ‘barely noticeable’ difference in glass colour if the products were side-by-side on the shelf some said they would ‘go for the flint one’. Finally, a number of participants said they were ‘more influenced by price/brand’ and that ‘cost could make a difference’ to their purchase decisions.


2.8.5.10 Gin Some participants preferred the flint perhaps because they were ‘more used to it’ in flint. One person said they thought of gin as a clear liquid and therefore wanted a ‘clear bottle’. Other comments included that it ‘looked stronger’ in the flint bottle and one gin drinker preferred flint bottle. Another said ‘if it’s cheap you need a clear bottle’. However, in contrast many said the gin ‘looked cheaper’ in the flint bottle, ‘looked like vodka’, because some gins are in green bottles, for example Gordon’s gin was mentioned many times by participants. The green bottle looked as if it ‘might be a bit more special’, ‘more expensive’ ‘more appetising’ ‘drier or stronger tasting’. In fact one said it would be ‘better in an even darker green bottle’. Some did not like the ‘generic’ bottle and felt gin should be in a ‘squarer’ bottle again perhaps reflecting the familiarity of Gordon’s. Finally, one person summed things up nicely by stating that ‘people are more accepting of coloured glass for liquids (as opposed to food products)’ and that they might be ‘more loyal to brand/product for food items as they have built up stronger habits with respect to food products’. 2.8.5.11 Comments on recycling and marketing of ‘green glass’ Comments here were indistinguishable as those made in Study Two. 2.9 Study Two – discussion Study Two replicated the methodology used in Study One with the exceptions that the green glass was a lighter colour and that an additional product, Rosé wine, was tested. As with Study One the analysis of the rating scales for all products showed a number of significant differences in favour of the flint glass. Of the ten rating scales six showed a significant difference. However, as with Study One, when the products were considered individually differences were apparent. In a similar vein to the findings from Study One there were significant effects of glass colour on most of the rating scales for mayonnaise. The Basics and Regular versions showed statistically significant effects on the same rating scales, specifically: Purchase likelihood, ‘I think this product is high quality’, ‘I think the contents of this bottle / jar are pure’, ‘I think the appearance of this bottle / jar’ is typical for a product of this type’, ‘I think the appearance of this bottle / jar is pleasing’, ‘I think the contents of this bottle / jar are appetising’ and ‘I think the contents of this bottle / jar are fresh’. In all cases the flint jar was rated higher than the green jar. Focus group comments were also consistent with this preference, and showed a general dislike for this product when presented in the light green glass. Comments included observations such as ‘it looks like it has gone off / mouldy’ or that it looked ‘radioactive’ or ‘toxic’. Others stated that the green glass would ‘put them off’ and that they wouldn’t pick it up. A few participants made vaguely positive comments such as ‘if I knew that the glass was recycled then I would go for it’ and ‘the green glass makes it look richer’. However, these comments were in the minority. Also consistent with the results of Study One was the interaction between gender and glass colour, with a preference for flint being stronger for women than men on the rating of quality. As was the case in Study One the majority of females (86%) reported that they did the shopping for their household compared to only 53% of males and this is likely to have implications for any change in glass colour on consumer behaviour given females clear dislike of this product. There were also two significant colour by age interactions for mayonnaise. For the Basic range version the 18-34 year old group considered the green jar to be more attention grabbing, although given the findings relating to purchase likelihood this is not a positive outcome. The 55 and over group also considered it to be slightly more attention grabbing whereas the 35-55 year old group considered it less attention grabbing. For the regular mayonnaise all age groups considered the green jar to contain a less fresh product but this effect was more marked for the 18-34 year olds than the other age groups. Overall, it is clear that consumers show a dislike for mayonnaise presented in green glass regardless of age or gender. Six products showed no significant main effects of the use of the lighter green recycled glass. These were Pinot Bianco, Chardonnay, rosé wine, Whisky and both ranges of jam. However, as with Study One it is essential to note that statistical tests are not designed to look for no differences and consequently the results must be interpreted with some caution. Again, interactive effects and focus group comments cast further light on the situation. There were two significant colour by gender interactions for Pinot Bianco. Males rated the green bottle as more attention grabbing than flint bottle, with a small difference in the opposite direction for females, and males rated the green bottle as more typical for the product than the flint bottle with an opposite difference for females. Focus group comments related to the two white wine products were broadly similar with a number of participants saying that they ‘did not notice the colour difference’ and ‘can’t see a problem with the green bottle’ or when they did notice the difference they stated that it would ‘not influence their purchase decisions’. The difference was said


to be more noticeable for Chardonnay than Pinot Bianco but as noted above, it seems this would not influence purchase decisions. On the basis of these results it seems that presenting these products in light green bottles is likely to have little effect on consumer behaviour. For the Rosé wine focus group comments were more polarised with some stating that they would be ‘equally happy with either’ and others stating that they preferred the flint as it looked ‘more summery and fresher’. Conversely other participants claimed that they preferred the light green bottle as it made the wine look more full-bodied’. These polarised comments make a firm conclusion difficult to draw. However, a comment worthy of note was that the rosé wine appeared to be a ‘dark variety’ and this may have had an additional influence on consumers’ purchase preferences. Neither of the Jam products showed any main effect of glass colour on any of the rating scales, although there were a number of significant interactions of both colour and age and colour and gender. For jam from the regular range there was a significant interaction of colour and age with 35-55 year olds rating the product as easier to evaluate in the green jar and the over 55 year olds rating the flint jar higher on this measure while there was no effect for the 18-34 year olds. This product also showed two significant colour by gender interactions with males rating jam in the green jar as less appetising than in the flint jar, with a small effect in the opposite direction for females. For the rating of freshness males rated the green jar as containing a less fresh product while females rated the green jar as containing the fresher product. For the jam from the Basic range there was a significant colour by gender interaction with males rating the purchase likelihood of jam in the green jar as lower than in the flint jar while there was little effect of glass colour on this rating for females. Focus group comments were in keeping with the lack of any statistical difference being found as participants claimed that they largely did not notice the difference, although some claimed that the difference in colour was more noticeable for the basic range version of this product. The whisky, showed no significant effect on any of the rating scales, but as with Study One, focus group comments were somewhat mixed with some claiming that the green bottle made the whisky ‘appear stronger’ and ‘more alcoholic’. However, there were a number of negative comments on the green glass with some noting that it looked less appealing and ‘cheaper’. Overall there is no clear picture here regarding the light green glass, but it looks to be a less positive outcome than some of the other alcohol products. There was one significant effect found for gin with participants rating the flint bottle as containing a purer product than the green bottle. There were no other significant differences. As with Study One, focus group comments were polarised. Some preferred the flint bottle as the contents ‘looked stronger’ or because it suited the clear liquid. Others preferred the light green bottle saying that the flint bottle ‘looked cheap’ whereas it appeared ‘stronger tasting’ and ‘more expensive’ in the light green. There was one significant main effect for pasta sauce with it being rated as easier to evaluate the product in the flint jar than the green jar. There were also significant colour by gender interactions for this product with males rating the green jar as making the product appear to be of lower quality than in the flint jar, with a small effect in the opposite direction for females. Also, males rated the green jar as making the product appear less appetising than the flint jar with no effect for females. Again demographic information regarding consumers may need to be considered in relation to this product. Overall comments from the focus group suggested that the effect of the light green glass on this product was largely negative, with the flint jar making the product appear more appetising and fresher and the green jar making the product look like it had been ‘left open for a while’.


3.0 Conclusion In this report we present the results of two studies that compared consumers’ responses to products in flint glass containers and green-tinted glass containers that were designed to represent the effects of increased use of recycled glass and/or decreased use of decolourisers. Study One compared flint glass with a relatively darker green glass. Study Two compared flint glass with a relatively lighter green glass. In this section of the report we draw together results of these studies and consider what conclusions can be drawn. Initially we do this for each product/product type. We then identify some more general design issues that influence colour-related perceptions. Finally we consider implications of the findings. Mayonnaise: considering the results of the two studies together, ratings data and the focus group comments consistently indicate that consumers consider the use of green glass for mayonnaise to be unacceptable. This response often persisted even when they were informed as to why the green glass is being used. Ratings of the product and of purchase likelihood suggest that consumers would avoid mayonnaise products in green glass. Interactions for some of the rating scales suggest that this effect may be stronger for women. Ratings and focus group comments indicated that consumers frequently attributed the perceived colour to the product (contents) rather than the glass (container). Jam: there were some differences in the results of Study One and Study Two. In Study One the ratings data indicated some (marginal) advantages for the green container for the Basics range (label). Interactive effects suggested that any preference, on average, was restricted to males, while the middle age group, on average, preferred the flint glass. There were no effects for the Regular range (label). For Study Two there were no significant main effects of glass colour. This may be attributable to these effects not being strong and the sample size for the second study being smaller. However, there were interactive effects of glass colour and gender indicating that males, on average, preferred the flint containers (this applied to both the Basics and the Regular ranges). This also contrasts with findings of Study One and also with the interactive effects of colour and gender found for mayonnaise. It is possible that differences in results for the two studies reflect differences in the green glass colour (the colour for Study Two was lighter than that for Study One). Comments from the focus groups indicated that the manipulation of glass colour altered consumers’ perceptions of the taste and quality of the product, with jam in the green containers being perceived as richer and perhaps stronger tasting, while jam in the flint container was regarded by some as fresher. It seems that the manipulation of glass colour was having differential effects on consumer perceptions along a number of relevant dimensions. Overall, the data suggest that glass colour does not have a strong influence on purchase preference, but this may be dependent on the exact colour of the container and the demographics of the consumer. Pasta Sauce: the effects of glass colour on ratings also differed between the two studies. For Study One ratings indicated a purchase preference for the flint container, and that participants thought the product was more appetising in this container. For Study Two the only significant main effect was that participants thought it easier to evaluate the product in the flint container. An interaction in Study One indicated a similar effect but limited to female participants. For Study Two an interaction between glass colour and gender also indicated that males found the product in the green glass less appetising. This is consistent with results for jam from this study (but see above). Comments from the focus groups, as for the jam, suggest that there may be some advantage for the green glass in making the product appear richer and stronger tasting (although whether this has a positive effect on purchase will depend on taste preference) but that the flint container makes the product appear fresher. Whisky: there were no significant main effects of the colour manipulation on ratings in either study. This must be qualified, however. Focus group responses indicated a polarisation of opinion, with some participants expressing a preference for the flint bottles and some expressing a preference for the green bottles. Although there was a good deal of individual variability in responses, the product was considered by some to be stronger and more alcoholic in the green, but often also to look cheaper. In the flint glass the product was regarded by many as richer and purer (especially in Study One). Related to this a significant interactive effect was found in Study One, such that the older age group were less likely to purchase the product in the green container. Following from this, effects of glass colour change are likely to be idiosyncratic, but predicting impact on purchase preferences would need to take account of consumer demographics and purchase motivations. Gin: there was just one significant main effects of the colour manipulation on ratings, with the product being rated as purer in the flint bottle in Study Two. However, as with the Whisky, focus group responses indicated polarisation of opinion. Comments varied, but the quality of the gin and the cheapness (appearance) of the product were key factors attributed to colour. The use of strong colour in the product of current market leaders was a factor that was mentioned with regard to acceptance. Significant interactive effects for Study One indicated age differences in colour-related perceptions of typicality. There was also a significant interaction between age


group and glass colour for ratings of product freshness, with ratings for the middle age group (35-55 years) being contrary to the other age groups, being higher for the flint. White wine: there (Chardonnay and Pinot Bianco) were no significant main effects of the colour manipulation on participants’ ratings. In Study One there was a significant interaction between age group and colour, such that the middle age group thought the product appeared stronger in the flint glass, whereas the other age groups thought it appeared stronger in the green. In Study Two there were also significant interactions between gender and bottle colour for ratings of typicality and, perhaps following from this, ratings of how attention grabbing the product would be (males thought the product less typical and more attention grabbing in the green; females vice versa). Focus group comments indicated that the colour difference was not very noticeable, but the flint glass made the product look lighter in taste and the green glass made the product look stronger in taste. Rosé wine: only examined in Study Two. There were no significant effects of glass colour. However, as with several of the other products, comments from the focus groups indicated divided opinion with regard to preference. The wine appeared more ‘full-bodied’ in the green, and lighter and fresher in the flint. An important issue that was raised in the focus groups concerns the relationship between the colour of the glass and the colour of the label and lid. For some products it was noted that the white lid and label (particularly those from the Basics range) emphasised the green colour of the glass. This was frequently considered off-putting. Related to this, the ‘fill level’ of products was also identified as an influential variable. The lower the fill level, the more obvious the green glass and this was noted by a number of participants. This indicates that any change in glass colour should be part of a broader consideration of the packaging design (see also previous report: WRAP report GLA0039 ‘Colourite project – maximising cullet additions in the glass container industry’, in which label colour was also identified as an important variable). When considering implications, the results of these studies suggest that for some products (e.g., mayonnaise) the use of green glass (of any noticeable shade) is likely to prove unacceptable to consumers. It may be the colour/nature of the product plays an important role in consumers’ perceptions and responses. For darker coloured food products and for alcoholic drink products the effects of container colour were less consistent. Ratings for purchase likelihood for pasta sauce were higher when presented in flint glass (Study One), but for many products there was no statistically significant difference, or even a small advantage associated with the use of a green container (Basics jam in Study One). However, caution must be exercised when interpreting these results for the following reasons. First, results of the focus groups indicated that the change in container colour influenced participants’ perceptions of the products along a number of dimensions, including taste, freshness, and price. Generally green glass was associated with a stronger, richer product taste and the flint glass was associated with a fresher, purer product. If consumers’ product preference is a function of product attribute priorities and product attribute values, colour-related differences in outcome can be predicted. Second, there was a good deal of inter-individual variability in the nature of these effects. For some products similar means for ratings seemed to mask more polarised between-participant views. Some of this was predictable in terms of the interactive effects of age or gender of participants (e.g., in Study One the older group gave lower purchase likelihood ratings to whisky when presented in green glass, but this was not the case for other age groups). Third, statistical tests are designed to identify the reliable presence of a difference between means rather than its absence. Given a larger sample and/or different methods statistically significant differences may be identified. These results should be taken as indicative rather than absolute. Finally, the focus groups identified that other design features, such as label colour, label size, cap colour, and fill level influence consumer perceptions of glass colour. Consideration of these factors should feature in any container changes. The design of competitor products will also be influential (e.g., many participants thought green gin bottles and wine bottles acceptable because of their prevalence in the marketplace). From this it would seem that any move towards the use of green glass containers for food and drink products would need to: i) be product specific; ii) consider all product design elements (e.g., label colour and size, lid, fill level); and, iii) consider consumer demographics (existing and anticipated customer base). 4.0 Demand from Retailers and Brand owners The demand from retailers and brand owners for glass containers with high recycled mixed glass content will be ascertained by a survey in October 2009 after publication of this report. The findings of the survey will feed into forthcoming research and development work to be funded by WRAP that will assess the practicalities and economics of producing containers from mixed colour cullet.


Appendix 1:pre-rating questionnaire and rating questionnaire Participant Number: ………….. Design ……………….. Order …………………. 1. Your age: …….years 2. Gender: Male / Female 3. What is your highest education level? Tick one box only. None Completed secondary/high school Completed College Completed University Completed a postgraduate degree Vocational 4. What is the current occupation of the highest earner of your household? Tick one box only. Higher managerial, administrative or professional Intermediate managerial, administrative or professional Supervisory or clerical and junior managerial, administrative or professional Skilled manual worker Semi and unskilled manual worker Retired, State pensioner or widow (not working), casual/lowest grade worker 5. What is your total household income? Tick one box only. Less than £15,000 £15,000 to £25,000 £25,000 to £35,000 £35,000 to £45,000 £45,000 to £55,000 £55,000 to £65,000 £65,000 to £75,000 More than £75,000 6. What is your ethnic background? Tick one box only. White British White Other Black British African African Caribbean Chinese Bengal Indian Arab Bangladeshi Pakistani Mixed Other (please specify)


7. Who usually does the shopping? You Other (please say who) 8. Which supermarket do you use to do your main shopping? Tick one box only. Aldi Asda/Walmart Lidl Morrisons Netto Sainsburys Somerfield Tesco Waitrose Other (please specify) 9. Which other supermarket(s) do you use for supplementary shopping? Tick all that apply. Aldi Asda/Walmart Lidl Morrisons Netto Sainsburys Somerfield Tesco Waitrose Other (please specify)


In this session today, we will firstly show you a number of food and drink products and ask you to rate each of them with respect to a number of adjectives. Some adjectives may seem a little unusual in this context; don’t worry. Work fairly quickly without puzzling over individual items for long periods. It is your first impression that we want. Sometimes you may feel as though you have seen the same product before. This will not be the case; every product is different. Do not try to remember how you marked similar products earlier in the session. Mark each product based on a separate and independent judgment. For each question circle the number between 1 and 7 that best represents your response. So for example, I think the appearance of this product is Practical :

Definitely not

Definitely

1 2 3 4 5 6 7


PRODUCT EVALUATED: _________________________________ 1. I would like to purchase this product:

Definitely Not

Definitely

1 2 3 4 5 6 7 2. I think this product is attention grabbing:

Definitely Not

Definitely

1 2 3 4 5 6 7 3. I think this product is high quality:

Definitely Not

Definitely

1 2 3 4 5 6 7 4. I think the contents of this bottle/jar are pure :

Definitely Not

Definitely

1 2 3 4 5 6 7 5. I think the appearance of this bottle/jar is typical for a product of this type:

Definitely Not

Definitely

1 2 3 4 5 6 7 6. I think the appearance of this bottle/jar is pleasing:

Definitely Not

Definitely

1 2 3 4 5 6 7 7. I think the contents of this bottle/jar are appetising:

Definitely Not

Definitely

1 2 3 4 5 6 7 8. I think the contents of this bottle/jar are fresh:

Definitely Not

Definitely

1 2 3 4 5 6 7


9. I think the contents of this bottle/jar will be strong tasting:

Definitely Not

Definitely

1 2 3 4 5 6 7 10. It is easy to tell whether I would like the contents of this bottle/jar:

Definitely Not

Definitely

1 2 3 4 5 6 7 11. How much do you think this product will cost? __________________________________ 12. Now that you have seen all the products, what do you think was the purpose of this experiment? ** N.B. This last item will only appear on the final rating questionnaire. **

Please check that you have answered all the questions.

Please wait for the next product. Do not turn over the page until instructed by the researcher.

Thank You.


Appendix 2: Products



www.wrap.org.uk/maximumrecycledcontent

Documents

Maximum Recycled Content Report 191009 - WRAP Recycled Content...glass. For whisky and gin, when considering the full sample, there was just one statistically significant difference