Experimentation in Market Reasearch

Embed Size (px)

Citation preview

  • 8/4/2019 Experimentation in Market Reasearch

    1/86

    Business Research Method

    Experimentation

  • 8/4/2019 Experimentation in Market Reasearch

    2/86

    An Introduction to Experimental Design

    Statistical studies can be classified as being either

    experimental or observational.

    In an experimental study, one or more factors(

    Independent Variables) are controlled so that data

    can be obtained about how the factors influencethe variables of interest(Dependent Variable).

    In an observational study, no attempt is made to

    control the factors. Cause-and-effect relationships are easier to

    establish in experimental studies than in

    observational studies.

  • 8/4/2019 Experimentation in Market Reasearch

    3/86

    Experiment

    . Experiment: A research method inwhich conditions are controlled so that

    one or more independent variables can

    be manipulated and their effect ondependent variable is measured .

    To study impact of a sports coach oncricket teams performance an

    experiment can be conducted

  • 8/4/2019 Experimentation in Market Reasearch

    4/86

    Experiment

    .-A research investigation in which conditionsare controlled

    -One independent variable is manipulated

    (sometimes more than one)-Its effect on a dependent variable is

    measured

    -To test a hypothesis

  • 8/4/2019 Experimentation in Market Reasearch

    5/86

    Some Definitions

    Absolute Experiment: If we want to know

    impact of a sports coach on cricket teamsperformance it is an absolute experiment

    Comparative experiment :If we want toknow impact of one sports coach on cricket

    teams performance compared to impact ofanother sports coach on cricket teamsperformance it is a comparative experiment

    Experimental units: Subjects or entitiessubjected to various treatments are calledexperimental units

  • 8/4/2019 Experimentation in Market Reasearch

    6/86

    Some Definitions

    - Experimental group: The group ofsubjects exposed

    to an experimental treatment.Group of students provided with a sport coach

    - Control Group :A group of subjects who are

    exposed to the control conditions in an experiment,that is they are subjects not exposed to the

    experimental treatment

    Group of students not provided with a sports coach

  • 8/4/2019 Experimentation in Market Reasearch

    7/86

    Treatment

    - Treatment: It refers toconditions to which

    experimental groups aresubjected and control groups

    are not subjected

    Provision of sports coach

  • 8/4/2019 Experimentation in Market Reasearch

    8/86

    Extraneous Variable

    Extraneous variable: Independent variables that

    are not directly linked with the study but mayinfluence dependent variable:

    If the study wants to test a hypothesis that a

    relationship exists between academicachievement and self concept of children then

    academic achievement is dependent and self

    concept is independent variable.

    Intelligence may also affect academic

    achievement but it is not related to studys

    purpose and hence it is an extraneous variable.

  • 8/4/2019 Experimentation in Market Reasearch

    9/86

    Experimental Design

    Experimental Design is a set of

    procedures specifying

    --Manipulation of the Independent

    Variable--Selection of Dependent Variable

    --Experimental groups & Control groups

    --Control Over Extraneous Variables

  • 8/4/2019 Experimentation in Market Reasearch

    10/86

    Manipulation of independent Variable

    The experimenter has some degree of control over the

    independent variable The variable is independent because its value can be

    manipulated by the experimenter.

    To test relationship between package design & sales

    Expose customers to three package designs A,B &C.These packs are placed on shelves of select outlets

    Consumers response is measured

    Package design is independent variable ,which is

    manipulated & there are 3 treatment levels A,B & C

  • 8/4/2019 Experimentation in Market Reasearch

    11/86

    Selection of Dependent Variable

    Selection of dependent variable is another important

    decision Sales volume of the product is considered as dependent

    variable in package design example

    Selecting dependent variable may not be easy in all

    cases

    To evaluate effectiveness of various ad programmes

    dependent variable can be brand image, brand

    awareness & product sales Purpose of the research helps in selection of

    appropriate dependent variable

  • 8/4/2019 Experimentation in Market Reasearch

    12/86

    Experimental & Control Groups Experimental group is a group of test units exposed to

    change in independent variable Control group is a group of test units not exposed to

    change in independent variable

    In package design example a group of supermarkets

    (Experimental group) is selected & each package designis displayed for a month

    Another group of supermarkets (Control group)continue to carry regular package design for that period

    Sales of product are measured for both groups &difference in sales is analysed to find out impact ofdesign on sales

  • 8/4/2019 Experimentation in Market Reasearch

    13/86

    Controlling Extraneous Variables

    Extraneous (Confounding)variables need to be controlled

    otherwise it is difficult to find out

    whether the change in dependent

    variable is due to change in

    independent variable or change inextraneous variable

  • 8/4/2019 Experimentation in Market Reasearch

    14/86

    Ways to control extraneous variable

    Randomization Hold Conditions Constant

    Matching Subjects

    Design Control

    Statistical Control

  • 8/4/2019 Experimentation in Market Reasearch

    15/86

    Randomisation Process of assigning test units randomly to

    experimental treatments and assigning

    experimental treatments randomly to test units It helps to spread effects of extraneous variable

    equally over the test units

    Most popular method

    Not effective for small samples----Department stores assigned randomly toExperimental & Control groups

    ----If 3 versions of test commercials are to be

    assigned to 3 groups then one commercial isselected at random & assigned to a randomlyselected group & process is repeated

  • 8/4/2019 Experimentation in Market Reasearch

    16/86

    Physical Control of Extraneous Variables

    Hold Conditions Constant: Thelevel of extraneous variables is

    kept constant throughout the

    experiment

    -- If price is the extraneous variable

    prices are not changed duringexperimental period

  • 8/4/2019 Experimentation in Market Reasearch

    17/86

    Physical Control of Extraneous Variables

    Matching Subjects: Judgmental sampling isused to assign test units to both

    experimental group & control group so that

    both groups are matched in terms of

    characteristics of test units

    ---Department stores are matched on the

    basis of annual sales, size and location. One

    store from each matched pair would beassigned to each experimental group &

    other to control group

  • 8/4/2019 Experimentation in Market Reasearch

    18/86

    Statistical Control

    Extraneous variables affecting dependent

    variable are identified & measured usingappropriate statistical tools like ANOVA. Then

    effects of extraneous variable on dependent

    variable are removed by statistical adjustments

    If 5 departmental stores are used to study salesof products before & after in- house sales

    promotion then impact of the characteristics of

    the stores can be removed through use of

    ANOVA-randomised block design using

    department stores as blocks

  • 8/4/2019 Experimentation in Market Reasearch

    19/86

    Design Control

    Selecting appropriate experimental

    designs to conduct experiments helps incontrolling specific extraneous variables .

    Experimental Designs are of 4 types

    -Pre Experimental-True Experimental

    -Quasi experimental

    -Statistical

  • 8/4/2019 Experimentation in Market Reasearch

    20/86

    Experimental Validity Validity: Extent to which a research process

    is accurate & reflects actual marketconditions

    Internal Validity: It measure the extent of

    change in dependent variable in anexperiment caused by manipulation ofindependent variable .( Not vitiated byextraneous variables)

    External Validity :It measures to what extentinferences derived from experiment can begeneralised to real environment

  • 8/4/2019 Experimentation in Market Reasearch

    21/86

    Internal Validity

    Control of extraneous variable is a

    necessary condition for establishing

    internal validity

    Internal validity is the basic

    minimum requirement of an

    experiment

    Six major types of extraneousvariables are threats to internal

    validity

  • 8/4/2019 Experimentation in Market Reasearch

    22/86

    Factors Influencing Internal

    Validity History

    Maturation

    Testing

    Instrumentation

    Selection

    Mortality

  • 8/4/2019 Experimentation in Market Reasearch

    23/86

    Type of Extraneous Variable Example

    History - Specific events in theenvironment between the Beforeand After measurement that arehistoric in nature (rare occurrence)

    Maturation - Subjects change

    during the course of the experimentdue to passage of time. Subjectsbecome tiredTesting - The Before measure alertsor sensitizes subject to nature of

    experiment or second measure.

    While test marketing

    new farm equipmentarea gets affected byfloodsImpact of promotional

    campaign for a car getsaffected by steep price risein petrol by Iraq warDrug trials carried over longeperiod of time results in som

    effect on account ofphysiological changesQuestionnaire about thetraditional role of women

    triggers enhanced awarenes

  • 8/4/2019 Experimentation in Market Reasearch

    24/86

    Instrument -Changes in

    instrument to minmise testingeffect result in response bias

    Selection

    test units selected forexperimental group differ fromthose assigned to control group

    Mortality - Sample attrition; somesubjects withdraw from experiment

    New questions about

    women are interpreteddifferently from earlierquestions.

    Heavy users in experimentagroup & moderate & lowusers in control group

    Subjects in TV viewing studdrop out before completion

  • 8/4/2019 Experimentation in Market Reasearch

    25/86

    External Validity

    Experiments conducted in natural

    settings offer more external

    validity as compared to

    experiments in controlledconditions.

    Since generalisation is easier with

    field conditions

    i i

  • 8/4/2019 Experimentation in Market Reasearch

    26/86

    Experimental Environment Laboratory Environment: Experiments conducted in controlled

    conditions

    ---Showing ads or products to select consumers in controlled

    conditions----blind taste tests

    Advantages-

    Effects of extraneous variables can be minimised

    -History effects eliminated

    -Provides more internal validity

    -time & cost effective

    -reduces the risk of information about products & ideas being passedon to competitors

    Disadvantages -Expts conducted in artificial conditions &results may not hold well

    in actual conditions

    -less external validity

    -results influenced by testing effects ,Test units aware of being tested

    & may not respond naturally

  • 8/4/2019 Experimentation in Market Reasearch

    27/86

    Field Environment Experiments conducted in natural settings

    --Launching products in select regions

    --Observing consumer behaviour regarding a POP display

    in supermarkets

    --analysing customer response to trial offers Advantages

    -high degree of external validity

    Disadvantages -Low degree of internal validity

    -Require greater time& efforts and are expensive

  • 8/4/2019 Experimentation in Market Reasearch

    28/86

    Laboratory Experiment Field Experiment

    Artificial-Low Realism

    Few Extraneous

    Variables

    High control

    Low Cost

    Short Duration

    Subjects Aware of

    Participation

    Natural-High Realism

    Many Extraneous

    Variables

    Low control

    High Cost

    Long Duration

    Subjects Unaware of

    Participation

  • 8/4/2019 Experimentation in Market Reasearch

    29/86

    Different Basic Experimental Designs

    Pre-Experimental Designs

    True Experimental Designs

    Quasi Experimental Designs

    Statistical Designs

    N i

  • 8/4/2019 Experimentation in Market Reasearch

    30/86

    Notations

    X: Treatment or manipulation of independent

    variable O: Observation or measurement of dependent

    variable

    R: Random assignment of test units to

    experimental groups

    EG : Experimental group exposed to experimental

    treatment

    CG :Control group not exposed to experimentaltreatment

    Notations in the diagram are read from left to right

  • 8/4/2019 Experimentation in Market Reasearch

    31/86

    Conventions

    Movement from left to right indicates

    movement through time

    Horizontal alignment of symbols implies

    that all those symbols refer to specific

    treatment group

    Vertical alignment of symbol implies that

    those symbols refer to activities or events

    that occur simultaneously

    Examples

  • 8/4/2019 Experimentation in Market Reasearch

    32/86

    Examples

    X O1 O2 means that a given group of test

    units was exposed to treatment variable X

    and the response was measured at two

    different points in time ,O1 and O2

    R X1 O1

    R X2 O2 means that two groups of

    test units were randomly assigned to two

    different treatment groups at the same time

    and the dependent variable was measured in

    two groups simultaneously

  • 8/4/2019 Experimentation in Market Reasearch

    33/86

    Pre-Experimental Designs

    ---They lack proper control mechanism to deal

    with effects of extraneous variables

    --- Randomization is absent

    One Shot Design (After Only)

    One Group Pretest-Posttest

    Static Group Design

  • 8/4/2019 Experimentation in Market Reasearch

    34/86

    One Shot Design (After Only)

    It involves exposing experimental group to treatment X

    after which the measurement (O1) of dependent variable istaken .

    EG : X O1

    To measure effectiveness of a test commercial

    TV programme containing test Ad (X) is selected

    Telephone interviews are conducted of respondents whoreport having watched the TV programme previous night

    Respondents are asked whether they recall having seen theAd. Measurements of Dependent variable are recalls (O1)

    Recalls are compared with norm scores for interpretation.

    One Shot Design (After Onl )

  • 8/4/2019 Experimentation in Market Reasearch

    35/86

    One Shot Design (After Only)

    Drawbacks

    No random assignment of test units Extraneous variables are not controlled

    Internal validity low

    There is no other measurement against

    which O1 can be compared (what would

    happen if X was absent)

    More appropriate for exploratory studiesand not conclusive studies

    O G P P

  • 8/4/2019 Experimentation in Market Reasearch

    36/86

    One Group Pretest -Posttest Exposing an experimental group to experimental treatment (X) and

    taking measurements of dependent variable before(O1) and after(O2)the treatment

    EG: O1 X O2

    Difference between O1 and O2 measures the impact of treatment ondependent variable

    To measure effectiveness of a test commercial

    Respondents are recruited on the basis of convenience Administered a personal interview to measure attitude towards the

    product

    Then they watch a TV programme containing Commercial

    After watching TV programme they are again administered a personal

    interview to measure attitude towards product Effectiveness of test commercial is measured as O2-O1

    Influenced by maturation,testing ,mortality, selection bias affecting IV

  • 8/4/2019 Experimentation in Market Reasearch

    37/86

    Static Group Design Two groups of test units are involved EG &CG

    EG is exposed to treatment CG is not exposed

    Measurements are taken for both groups after Expt.

    EG: X O1

    CG: O2

    Difference between O1& O2 is effect of treatment To test effectiveness of TV commercial

    Two groups of respondents are recruited

    EG is exposed to TV programme containing Ad

    Attitude towards product of both EG &CG are measured Effectiveness of test Ad is measured as O1-O2

    Selection bias ,mortality affects results with less IV

  • 8/4/2019 Experimentation in Market Reasearch

    38/86

    True Experimental Designs

    ---Use randomisation & one or more controlgroups to reduce effect of extraneous variables

    Pre testPost test Control Group Design

    Posttest only Control Group DesignSolomon Four Group design

    Pretest-Posttest Control Group Design

  • 8/4/2019 Experimentation in Market Reasearch

    39/86

    Pretest Posttest Control Group Design

    Two groups of test units EG &CG are considered

    Test units are assigned to these 2 groups randomly Pre test measurements of dependent variable are

    taken for both groups

    EG is exposed to treatment Post test measurements of dependent variable aretaken for both groups

    EG: R O1 X O2

    CG: R O3 O4

    Treatment Effect (TE) is (O2-O1)-(O4-O3)

    Pre Test Post Test Control Group Design

  • 8/4/2019 Experimentation in Market Reasearch

    40/86

    Pre Test Post Test Control Group Design

    To measure effectiveness of a TV commercial

    Sample of respondents is selected at random

    Half of these would be randomly assigned to EG & other halfwould form CG

    Both groups would be administered a questionnaire to obtain pre-test measurement on attitude towards the product

    Only EG would be exposed to TV programme containingcommercial

    Then respondents in both groups would be administered aquestionnaire to obtain post test measurements on attitude towardsproduct

    Effectiveness measure is (O2-O1)-(O4-O3)

    This design provides accurate results & addresses most ofextraneous variables accept testing effects

    P tt t t t l C t l G D i

  • 8/4/2019 Experimentation in Market Reasearch

    41/86

    Posttest test only Control Group Design

    Both EG & CG are formed by random assignment

    of test units EG is exposed to experimental treatment CG is not

    Post test measurement of dependent variable is

    taken for both the groups EG: R X O1

    CG: R O2

    Treatment Effect (TE) equals O1-O2

    P O l C l G D i

  • 8/4/2019 Experimentation in Market Reasearch

    42/86

    Posttest Only Control Group Design

    To measure effectiveness of a TV commercial

    Sample of respondents is selected at random Half of these would be randomly assigned to EG & other

    half would form CG

    Only EG would be exposed to TV programme containing

    commercial Then respondents in both groups would be administered a

    questionnaire to obtain post test measurements on attitudetowards product

    Difference in attitudes of EG &CG is used aseffectiveness measure

    P O l C l G D i

  • 8/4/2019 Experimentation in Market Reasearch

    43/86

    Posttest Only Control Group Design

    Involves only 2 groups & one measurement

    Significant advantage in terms of time ,cost& sample size

    Selection bias & mortality which are the

    major drawbacks can be controlled throughcarefully designed experimental procedures

    Because of its simplicity it is most popular

    in marketing design

    Solomon Four Group Design

  • 8/4/2019 Experimentation in Market Reasearch

    44/86

    Solomon Four Group Design It involves 4 groups ,2 EGs & 2 CGs

    Six measurements are taken 2 pretest &4 posttest

    Also known as 4 group 6 study design EG: R O1 X O2

    CG: R O3 O4

    EG: R X O5

    CG: R O6

    Design provides various observations which can beanalysed

    O2-O1; O4-O3; O2-O4; O5-O6

    Addresses all extraneous variables but is expensive& consumes time & efforts

    Q i E i t l D i

  • 8/4/2019 Experimentation in Market Reasearch

    45/86

    Quasi Experimental Designs

    Are used when randomisation is not possible & timing oftest presentation lacks control of researcher

    Not as effective as true designs but better than pre-design

    Prominent is Time Series Design

    A series of measurements are taken on dependent variablefor a group of test units before & after the experimentaltreatment

    O1,O2,O3 X O4, O5, O6

    Trends before treatment are compared with trends aftertreatment to determine effectiveness of treatment

    Aids in identifying permanent change & temporary change

    Threats on account of history & instrumentation

    Quasi Experimental Designs

  • 8/4/2019 Experimentation in Market Reasearch

    46/86

    Quasi Experimental Designs

    To test effectiveness of a test commercial

    Broadcasting test commercial a predetermined number oftimes & examining data from a pre-existing test panel

    Scheduling of test commercial can be controlled but it isuncertain when or whether panel members are exposed toit

    Purchases of panel members before ,during and after thecampaign are examined to find out whether testcommercial has a short term effect ,long term effect or noeffect

    To find out impact of price change on sales of a product, aseries of observations are taken before the price change&after the price change. Trends before treatment arecompared with trends after treatment to determine change

  • 8/4/2019 Experimentation in Market Reasearch

    47/86

    Multiple Time Series Design

    Similar to time series except that another group

    of test units is added to serve as control group EG:O1,O2,O3 X O4,O5,O6

    CG:O7,O8,O9 O10,O11,O12

    Test commercial would be shown only in a few

    of test cities

    Panel members in these cities would form EG

    Panel members in the cities where commercial

    was not shown would constitute control group

  • 8/4/2019 Experimentation in Market Reasearch

    48/86

    One-Shot Design

    Internal Validity Problems

    History

    weak

    Maturation weak

    Testing

    not relevant

    Instrumentation

    not relevant

    Selection weak

    Mortality

    weak

  • 8/4/2019 Experimentation in Market Reasearch

    49/86

    One-Group Pretest-Posttest

    Internal Validity Problems

    History

    weak

    Maturation weak

    Testing

    weak

    Instrumentation

    weak

    Selection controlled

    Mortality

    controlled

  • 8/4/2019 Experimentation in Market Reasearch

    50/86

    Static-Group Design

    Internal Validity Problems

    History

    controlled

    Maturation possible source of

    concern

    Testing

    controlled

    Instrumentation

    controlled

    Selection weak

    Mortality

    weak

  • 8/4/2019 Experimentation in Market Reasearch

    51/86

    Pretest-Posttest Control

    Internal Validity Problems History

    controlled

    Maturation controlled

    Testing

    controlled

    Instrumentation

    controlled

    Selection controlled

    Mortality

    controlled

  • 8/4/2019 Experimentation in Market Reasearch

    52/86

    Posttest-Only Control

    Internal Validity Problems

    History

    controlled

    Maturation controlled

    Testing

    controlled

    Instrumentation

    controlled

    Selection controlled

    Mortality

    controlled

  • 8/4/2019 Experimentation in Market Reasearch

    53/86

    Solomon Four-Group Design

    Internal Validity Problems History

    controlled

    Maturation controlled

    Testing

    controlled

    Instrumentation

    controlled

    Selection controlled

    Mortality

    controlled

    Statistical Designs

  • 8/4/2019 Experimentation in Market Reasearch

    54/86

    Statistical Designs

    Aids in measuring effect of more than one

    independent variable Helps in isolating effects of most

    extraneous variables

    Four prominent designs in this category

    -Completely Randomised

    -Randomised Block

    -Latin Square

    -Factorial

  • 8/4/2019 Experimentation in Market Reasearch

    55/86

    Advanced Experimental Designs are

    More Complex

    Completely randomized

    Randomized block design

    Latin square

    Factorial

  • 8/4/2019 Experimentation in Market Reasearch

    56/86

    Completely Randomized Design

    An experimental design that uses a random process

    to assign subjects (test units) and treatments toinvestigate the effects ofonly one independent

    variable (factor).

    There are n test units & ktreatments (levels offactor)

    The n test units are assigned to k treatments

    randomly. Post test measurements are evaluated

  • 8/4/2019 Experimentation in Market Reasearch

    57/86

    Completely Randomized Design

    3 test cities are randomly exposed to 3 test

    commercials

    Then measurements are taken on sales in these 3

    cities

    Differences in sales in 3 cities are analyzed to seewhich AD is effective

    Example: Home Products, Inc.

  • 8/4/2019 Experimentation in Market Reasearch

    58/86

    Example: Home Products, Inc.

    Home Products, Inc. is considering marketing along-lasting car wax. Three different waxes (Type

    1, Type 2,and Type 3) have been developed.

    In order to test the durability of these waxes, 5new cars were waxed with Type 1, 5 with Type 2,

    and 5 with Type 3. Each car was then repeatedlyrun through an automatic carwash until the waxcoating showed signs of deterioration. Thenumber of times each car went through the

    carwash is shown on the next slide.Home Products, Inc. must decide which wax to

    market. Are the three waxes equally effective?

    Example: Home Products, Inc.

  • 8/4/2019 Experimentation in Market Reasearch

    59/86

    p

    WAX WAX WAX

    Observation Type 1 Type 2 Type 3

    1 48 73 51

    2 54 63 63

    3 57 66 61

    4 54 64 54

    5 62 74 56

    Sample Mean 55 68 57Sample Variance 26.0 26.5 24.5

    Example: Home Products, Inc.

  • 8/4/2019 Experimentation in Market Reasearch

    60/86

    Completely Randomized Design

    ANOVA Table

    Source of Sum of Degrees of Mean

    Variation Squares Freedom Squares F

    Treatments 490 2 245 9.55

    Error 308 12 25.667

    Total 798 14

    Example: Home Products, Inc.

    E l E t Oil C

  • 8/4/2019 Experimentation in Market Reasearch

    61/86

    Example: Eastern Oil Co. Completely Randomized Design

    Rejection RuleAssuming = .05, F.05 = 3.88 (2 d.f. numerator

    and 12 d.f. denominator). RejectH0 ifF>

    3.88.

    Test Statistic

    F= MSTR/MSE = 245/25.67= 9.55

    Conclusion

    Since 9.55 > 3.88, we rejectH0 and conclude

    that the miles per gallon ratings differ for the

    three gasoline blends.

    Completely Randomized Design

  • 8/4/2019 Experimentation in Market Reasearch

    62/86

    Completely Randomized Design To evaluate efficacy of a weight loss drug

    A sample of 40 consumers is selected

    Assigned randomly to two treatment levels 25 to treatment 1 & 15 to treatment 2

    Consumers in group 1 are asked to take drug for onemonth

    Consumers in group 2 are not given any drug After experiment measurements are taken for both groups

    & differences if any are analysed to find out effectivenessof the drug

    Applicable when test units are homogeneous Extraneous variables can be controlled

    A single variable is evaluated

  • 8/4/2019 Experimentation in Market Reasearch

    63/86

    Randomized Block Design

    An extension of the completely randomized

    design in which a single extraneous variablethat might affect test units response to the

    treatment has been identified and the effects

    of this variable are isolated by blocking outits effects.

    Example Randomised Block Design

  • 8/4/2019 Experimentation in Market Reasearch

    64/86

    Example Randomised Block Design

    A consumer product company wants to examine

    impact of price change on sales of its newly

    developed health drink

    Store type is extraneous variable which would also

    affect product sales

    Company has decided to conduct experiment using 3price levels & 3 store types

    Nine retail outlets are segregated according to store

    type Treatment levels (prices) are assigned randomly to

    each test unit ( retail Outlet)

  • 8/4/2019 Experimentation in Market Reasearch

    65/86

    Example Randomised Block Design

    Price

    Change

    Drug Store Supermarkets Malls

    Rs 10

    Rs 20

    Rs 30

  • 8/4/2019 Experimentation in Market Reasearch

    66/86

    E l E t Oil C

  • 8/4/2019 Experimentation in Market Reasearch

    67/86

    Example: Eastern Oil Co.

    utomobile Type of Gasoline (Treatment) Blocks

    (Block) Blend X Blend Y Blend Z Means

    1 31 30 30 30.333

    2 30 29 29 29.333

    3 29 29 28 28.667

    4 33 31 29 31.000

    5 26 25 26 25.667

    reatment

    Means 29.8 28.8 28.4

    The ANOVA Procedure

  • 8/4/2019 Experimentation in Market Reasearch

    68/86

    he ANOVA procedure for the randomized block desig

    equires us to partition the sum of squares total (SST) in

    hree groups: sum of squares due to treatments, sum of

    quares due to blocks, and sum of squares due to error.

    he formula for this partitioning is

    SST = SSTR + SSBL + SSE

    he total degrees of freedom, n - 1, are partitioned suchhat k- 1 degrees of freedom go to treatments,

    - 1 go to blocks, and (k- 1)(b - 1) go to the error term.

    The ANOVA Procedure

    Example: Eastern Oil Co

  • 8/4/2019 Experimentation in Market Reasearch

    69/86

    Example: Eastern Oil Co.

    Randomized Block Design

    Mean Square Due to TreatmentsThe overall sample mean is 29. Thus,

    SSTR = 5[(29.8 - 29)2 + (28.8 - 29)2 + (28.4 - 29)2] = 5.2

    MSTR = 5.2/(3 - 1) = 2.6 Mean Square Due to Blocks

    SSBL = 3[(30.333 - 29)2 + . . . + (25.667 - 29)2] = 51.33

    MSBL = 51.33/(5 - 1) = 12.8

    Mean Square Due to Error

    SSE = 62 - 5.2 - 51.33 = 5.47

    MSE = 5.47/[(3 - 1)(5 - 1)] = .68

    Example: Home Products, Inc.

  • 8/4/2019 Experimentation in Market Reasearch

    70/86

    Randomized Block Design

    ANOVA Table

    Source of Sum of Degrees of Mean

    Variation Squares Freedom Squares F

    Treatments 5.2 2 2.6 3.82

    Block 51.3 4 12.8

    Error 5.5 8 0.68

    Total 62 14

    p

  • 8/4/2019 Experimentation in Market Reasearch

    71/86

  • 8/4/2019 Experimentation in Market Reasearch

    72/86

    Factorial Design

    An experiment that investigates the

    interaction of two or more independentvariables on a single dependent variable.

  • 8/4/2019 Experimentation in Market Reasearch

    73/86

    F i l D i

  • 8/4/2019 Experimentation in Market Reasearch

    74/86

    Factorial Design Researcher is investigating believability of Ads on 0-100

    scale Two magazine Ads A & B are to be compared

    Gender of the reader is another factor

    This is a 2x2 factorial experiment Permits to test 3 hypothesis

    Which Ad is more believable (Main)

    Which gender tends to believe magazine Ads more(Main)

    Which gender finds which ad more

    believable(Interaction)

  • 8/4/2019 Experimentation in Market Reasearch

    75/86

    Men

    Women

    Ad A Ad B

    65

    65

    70 60

    Main Effectsof Gender

    Main Effects of Ad

    >

    2 x 2 Factorial Design

  • 8/4/2019 Experimentation in Market Reasearch

    76/86

    Price Red Gold

    $25 Cell 1 Cell 4$30 Cell 2 Cell 5$35 Cell 3 Cell 6

    Package Design

    Factorial Design -- Roller Skates

  • 8/4/2019 Experimentation in Market Reasearch

    77/86

    Latin Square Design

    A balanced, two-way classification scheme

    that attempts to control or block out the

    effect oftwo extraneous factors byrestricting randomization with respect to the

    row and column effects.

    Latin Square Example

  • 8/4/2019 Experimentation in Market Reasearch

    78/86

    To find out impact of three different Ads on sales

    Pricing & income levels of consumers are EV

    In latin square levels of all the 3 variables must be same (2EV and 1 INDEP)

    A table is then developed with levels of one EVrepresenting rows ,levels of second EV representing

    columns . Levels of independent variable (treatments) are exposed toeach cell on a random basis so that each row has alltreatments & each column has all treatments

    Treatment effect is determined for each cell

    Analysis shows which treatment level influences dependentvariable more

  • 8/4/2019 Experimentation in Market Reasearch

    79/86

    Latin Square Example

    Pricing

    Levels

    Low

    Income

    Middle

    Income

    High

    Income

    Rs 10000 Ad-B Ad-A Ad-C

    Rs 12000 Ad-C Ad-B Ad-A

    Rs 14000 Ad-A Ad-C Ad-B

    L ti S

  • 8/4/2019 Experimentation in Market Reasearch

    80/86

    Latin Square

    A, B & C identify three treatments

    Rows & Columns of the table identify

    confounding factors (extraneous variables)

    A taste test may be confounded by order of testing

    A taste test may also be confounded by individual

    test preference

    To control for these 2 factors each subject is

    exposed to every treatment & order in whichsubjects test is randomised.

  • 8/4/2019 Experimentation in Market Reasearch

    81/86

    1 2 3

    1 A B C2 B C A3 C A B

    Order of Usage

    SU

    BJECT

    Latin Square

  • 8/4/2019 Experimentation in Market Reasearch

    82/86

    Latin Square

    Interaction effects are assumed to be minimal or

    absent

    Number of treatment levels for both confounding

    factors must be equal

    Suppose a retail grocery chain wishes to controlfor shelf space & city where product is sold

    It markets in only 3 cities & wishes to experiment

    with 4 levels of shelf space

    Having unequal number of levels for each factor

    will eliminate latin square as a possibility

  • 8/4/2019 Experimentation in Market Reasearch

    83/86

    TEST MARKETING

    Controlled experimentation

    Not just tryingsomething

    out

    But scientifictesting

  • 8/4/2019 Experimentation in Market Reasearch

    84/86

    Test Marketing

    An experimental procedure that provides an

    opportunity to test a new product or a new

    marketing plan under realistic marketconditions to measure sales or profit

    potential.

  • 8/4/2019 Experimentation in Market Reasearch

    85/86

    Selecting a Test Market

    Population size

    Demographic composition

    Lifestyle considerations

    Competitive situation

    Media

    Self-contained trading area

    Secrecy

    C l M h d f T

  • 8/4/2019 Experimentation in Market Reasearch

    86/86

    Control Method of Test

    Marketing Small city Low chance of being detected

    Distribution is forced (guaranteed)