8
The 4 most common mistakes in market research September 2010

The 4 Most Common Mistakes in Market Research

  • Upload
    palio

  • View
    3.105

  • Download
    1

Embed Size (px)

Citation preview

Page 1: The 4 Most Common Mistakes in Market Research

The 4 most common mistakes in market research

September 2010

Page 2: The 4 Most Common Mistakes in Market Research

Mistake #1:Making generalizations about a population from a small sample

• Small sample sizes (<50) have a high degree of variability– More likely that inappropriate conclusions will be drawn about the population

• Large sample sizes (>200) are less variable– More likely that appropriate conclusions will be drawn about the population

Source: Hague P. A Practical Guide to Market Research. B2B International. 2007.

Page 3: The 4 Most Common Mistakes in Market Research

Small samples are best used for …

1. Generating a wide range of ideas– A small sample is capable of generating a range of ideas– With larger sample sizes, ideas will simply begin to be duplicated

2. Drawing “All” or “None” conclusions– If each member of a small sample says the same thing, and there was no

sampling bias, the conclusion is likely to be valid

3. Drawing “Some” conclusions– If testing an “All” or “None” conclusion, a small sample may be enough to

generate a single negative counterexample

4. Providing a gut check– Small sample sizes can mirror previously established empirical generalizations,

in which case some conclusions about the small group may be justified

Source: Bock T, Sergeant J. Small sample market research. International Journal of Market Research. 2002;44(2):235-244.

Page 4: The 4 Most Common Mistakes in Market Research

Mistake #2:Recruiting participants without regard for the desired conclusions

Do you plan to draw conclusions

about…

The larger population?

Yes No

Segments of the

population?

Yes

Recruit a very large random sample of the larger population where each

segment has at least 50 participants

Recruit random samples from each segment of more than 50

participants

NoRecruit a random sample of the

larger population of more than 50 participants

Recruit a targeted sample, focusing on characteristics of interest

(exploratory research)

• Always consider potential conclusions before recruiting participants– Failure to do so can result in samples of insufficient sizes to make appropriate

conclusions (see Mistake #1)

Source: Hague P. A Practical Guide to Market Research. B2B International. 2007.

Page 5: The 4 Most Common Mistakes in Market Research

Mistake #3:Selecting a method that poorly addresses the research objective

• Qualitative methods are best suited for exploration and development– Especially hypothesis development and in-depth probing

• Quantitative methods are best suited for evaluative research– Especially ratings, rankings, and other metrics

Objective Style Appropriate MethodTo identify problems or

opportunitiesExploratory

Brainstorming, qualitative research (IDIs or focus groups)

To uncover the rationale behind perspectives and

opinionsDiagnosis

Additional qualitative research (IDIs or focus groups)

To evaluate the validity of perspectives and opinions

Testing/DescriptiveQuantitative research (surveys or large, well-controlled IDIs or focus

groups)

Source: Hague P. A Practical Guide to Market Research. B2B International. 2007.

Page 6: The 4 Most Common Mistakes in Market Research

Mistake #4:Performing analyses without regarding the type of data

• Common errors in analysis– Performing averages on rankings or ordinal scales

• The distance between units in an ordinal scale is not consistent– For example, the difference between responding “agree” and “strongly agree” may not be the

same as the difference between “neither agree nor disagree” and “agree”

– Making comparisons on rankings, interval, or ordinal scales• Ex: 80°F is not twice as hot as 40°F

– 0°F does not represent zero heat

Data Type Description Example

NominalValues that are simply named and have an

arbitrary orderGeographic regions

OrdinalValues of a scale that can be placed in

orderRankings, some rating

scales

IntervalValues of a scale that can be placed in

order and whose units are equally spaced

Fahrenheit and Celsius, some rating

scales

RatioValues of a scale that have an order, units

of equal spacing, and a true zero valueHeight

Source: Cohen BH. Explaining Psychological Statistics. 2nd ed. New York, NY: John Wiley & Sons, Inc.; 2001.

Page 7: The 4 Most Common Mistakes in Market Research

How to avoid mistakes

• Increase efficiency– Prevent minor errors that can “snowball” into major problems– Draw stronger conclusions by comparing new findings to past research

• Increase quality– Research design and methodology have a major impact on the quality of research

• Identify the best research method

• Consider the sample size

• Do not underestimate the importance of the language in discussion guides and surveys

• Choose metrics that allow the audience to draw accurate conclusions from the data

• Engage and challenge researchers during each step of the process– Do not take methodology and research design for granted– Ask researchers to explain their methods, designs, and analyses

Source: Hague P. A Practical Guide to Market Research. B2B International. 2007.

Page 8: The 4 Most Common Mistakes in Market Research

If you’d like to find out more or have some research thoughts of your own to share, contact –

Carl Turner

VP, Research & Analytics - [email protected]

or

Bob Mason

SVP, Brand Strategy - [email protected]