Marketing Research
Computer Lab 2
Dr Euejung Hwang
[Link]@[Link]
Overview
• Today :
• Composite measure
• Re-coding variables
• Reliability test
• Pearson Correlation
• One sample t-test
• Independent samples of t-test
• Paired samples t-test
2
Composite Measure
o Transform > Compute Variables
• Composite measure
Composite Measure
o Transform > Compute Variables
• Composite measure
: Create a variable ‘Attitude’ by taking a mean
value from three items (i.e., Att1, Att2, Att3).
Add a name for your
composite measure
i) You can type in
‘mean’ (and
brackets) and
move each item
from the box in
the left.
OR
i) Choose a ‘mean’
function from the
list on the right
and do the same
Re-coding variables
o Transform > Recode into different variables
• How to recode reverse scored items.
E.g., Health consciousness items
Please respond to the following questions based on the extent to which you
agree or disagree with the statements (1: Strongly disagree; 7: Strongly agree).
1) I eat healthy foods.
2) I do not want to ask myself all the time, whether the things I eat are good
for me
Re-coding variables
o Transform > Recode into different variables
i) Move a variable
that you want to
recode to
‘Numeric Variable
→ Output
Variable’ box
ii) Put a new name
for the variable in
the ‘Output
Variable’ box,
then click
‘Change’
iii) Then, click ‘Old
and New Values’
Re-coding variables
o Transform > Recode into different variables
i) Put an original value
under ‘value’ in ‘Old
Value’ box and a new
value in ‘New Value’
box. Then Click ‘Add’
ii) You will need to do this
for all the values on the
scale (e.g., 2 → 6, 3 → 5,
4 → 4, 5 → 3, 6 → 2, 7 →
1) (See next page).
iii) Click ‘Continue’. If you
go back to the data,
you will see a new
variable has been
created.
Re-coding variables
o Transform > Recode into different variables
Reliability Test
o Analyze > Scale > Reliability Analysis
• Cronbach’s alpha
Reliability Test
o Analyze > Scale > Reliability Analysis
• Cronbach’s alpha
Reliability Test
o Analyze > Scale > Reliability Analysis
• Cronbach’s alpha
Inferential Statistics for Association
1. Chi-square test
2. Pearson Correlation
Pearson’s Correlation
o The Pearson’s Correlation Coefficient (r)
• A measure of the direction and strength of
association that exist between two variables
measured on at least an interval scale.
• Represents how linearly the variables are related.
Pearson’s Correlation
o Example
• To see whether product cuteness is positively
associated with the product attitude.
• Null hypothesis (Ho)
: Product cuteness and product attitude are
not related.
• Alternative hypothesis (Ha)
: Product cuteness and product attitude are
positively related.
Pearson’s Correlation
o Analyze > Correlate > Bivariate
Pearson’s Correlation
o Analyze > Correlate > Bivariate
Make sure you use a composite measure of ‘Attitude’ you created.
Pearson’s Correlation
o SPSS output and reporting
• The correlation between two variables is positive
and significant (r = .39, p < .001).
Inferential Statistics for Differences
o T-Test
• A t-test is an inferential statistic used to determine
if there is a statistically significant difference
between the means of two variables.
✓ One sample t-test
✓ Independent samples t-test
✓ Paired samples t-test
One sample t-test
o Example
: How familiar are you with Starbucks?
(1 = not at all; 7 = very much)
One sample t-test
o Analyze > Compare Means > One Sample T Test
One sample t-test
o Analyze > Compare Means > One Sample T Test
One sample t-test
o How familiar are you with Starbucks?
(1 = not at all to 7 = very much)
• Participants were familiar with Starbucks (t(154) =
32.57, p < .001; midpoint = 4).
Independent samples t-test
o Testing for Differences
• A statistical test that compares the means of
two independent groups to determine whether
there is a statistically significant difference
between the means in two unrelated groups.
Cute Ice Cream Scoop Neutral Ice Cream Scoop
Independent samples t-test:
Levene’s test
o Assumption #5
• There needs to be homogeneity of variances.
o Levene’s test
: The Levene’s test is used to assess the equality of
variances between to or more groups.
• Ho: Groups have equal variances
• Ha: Groups have different variances
Independent samples t-test
o Analyze > Compare Means > Independent
Samples T test
Independent samples t-test
o Analyze > Compare Means > Independent
Samples T test
Independent samples t-test
o SPSS output and reporting
• Levene’s test significance may be interpreted at p < .01
or p < .001.
• As long as n > 30 and n1≈ n2, t=test is robust to violations
of homogeneity of variance.
Independent samples t-test
o SPSS output and reporting
• An independent sample t-test was conducted to determine
whether the cute ice cream scoop was perceived to be cuter than
the neutral ice cream scoop. There was a significant difference in
the scores for neutral scoop (Mneutral = 5.05) and cute scoop (Mcute =
5.83) conditions (t(140) = 2.92, p < .01), indicating that cute (vs.
neutral) ice cream scoop was perceived to be cuter.
Inferential Statistics for Differences
1. One sample t-test
2. Independent samples t-test
3. Paired-samples t-test
Paired samples t-test
o Testing for Differences
• A statistical test used to compare the means of
two measurements taken from the same
individual.
• E.g., Exposure to scarcity information leads to
higher new food evaluation.
DV1 = Pre-exposure DV2 = Post-exposure
taste rating taste rating
Paired samples t-test
o Analyze > Compare Means > Paired-
Samples T Test
Paired samples t-test
o Analyze > Compare Means > Paired-
Samples T Test
You will need to create a composite measure for Pre-exposure taste
variable (measured using Pre_Att1, Pre_Att2, Pre_Att3) and Post-
exposure taste variable ((measured using Post_Att1, Post_Att2,
Post_Att3)
Paired samples t-test
o How much did you like the taste of the snack?
(1 = not at all to 7 = very much)
• There was a statistically significant improvement in
perceived taste after (vs. before) the exposure to
the scarcity information (t(21) = 2.48, p < .05).