NAME: CHRISTIAN JOHN A.
SALUDAR
MIDTERM ASSESSMENT #4
INSTRUCTIONS: Answer the following questions/problems. Late submission of the answers will
mean deduction of points.
1. Differentiate between crosstabs procedure and means command in SPSS. Give
examples.
CROSSTABS: Definition
a. Crosstabulation is a method by which one analyzes the relationship between
multiple variables. Unlike frequency which only tells the number of cases falling
into categories, cross tabulation allows one to determine the number of cases
falling under crossed variables.
b. Cross Tabulation can be used for continuous variables if and only if they are in
separate categories. An example of which is on scores(0-10, 11-20, etc.). This is
only used for purpose of data description and display.
c. In cross tabulation, categories of one variable determine the rows, while
categories of another variable determine the columns. The frequency in between
shows the frequency of that combination.
d. Also, cross-tabulation offers a wide array of options:
i. Cell Button
a. Observed Count – Actual number of subjects within
each cell
b. Expected Count – The expected value for each cell
c. Row Percentage – Percent of values in each cell for
that row
d. Column Percentage – Percent of values in each cell
for that column
e. Total Percentages – The percent of values in each
cell for the whole table
f. Unstandardized residuals – Observed value minus
expected values
ii. Statistics
a. One can select whether chi-square or correlations,
nominal(contingency coefficient, phi, and cramers v,
lambda, uncertainty coefficient), ordinal(gamma,
Somer's d, Kendall's tau-b, Kendall's tau-c), nominal
by interval(eta), kappa, risk, McNemar, and Cochran’s
and Mantel-Haenszel statistics
e. Example is crossing the variables gender(male & female) and ethnicity(Asian,
Black, White, and Hispanic). The CrossTabs command show the frequency of
crossed variables.
i. Male Asian
ii. Male Black
iii. Male White
iv. Male Hispanic
v. Female Asian
vi. Female Black
vii. Female White
viii. Female Hispanic
f. In addition, all other values selected in the options menu and statistics menu of
the crosstabs command will appear in the output window.
g. An example of a crosstabs output is presented below:
PROCEDURE FOR CROSSTABS IS STATED BELOW:
h. Click on Analyze (in Menu Bar) -> Descriptive Statistics -> Crosstabs
i. In the new window, one can do the following processes:
i. Select the desired variable for analysis
1. Click the variable on the left box and choose whether it shall
be part of the rows or columns (by clicking on the arrow next
to the corresponding box)
2. If one desires to cross-tabulate three or more variables, one
can select the variable and assign it to the last box (the
previous or next button are used to determine the order of
layer variables)
ii. Click on the desired buttons
1. To start from scratch(after an initial analysis was made),
click on reset.
2. To obtain information on the process, click on help
3. To cancel the process, click on cancel
iii. Click on the cell button
1. A new window will appear where one can determine the
counts(observed, expected, and hide small counts),
compare column properties, percentage(row, column, total),
noninteger weights (round cell counts, truncate cell counts,
no adjustments, round case weights, truncate case weights)
a. Observed Count – Actual number of subjects within
each cell
b. Expected Count – The expected value for each cell
c. Row Percentage – Percent of values in each cell for
that row
d. Column Percentage – Percent of values in each cell
for that column
e. Total Percentages – The percent of values in each
cell for the whole table
f. Unstandardized residuals – Observed value minus
expected values
iv. Click on statistics (to perform chi-square analysis)
a. One can select whether chi-square or correlations,
nominal(contingency coefficient, phi, and cramers v,
lambda, uncertainty coefficient), ordinal(gamma,
Somer's d, Kendall's tau-b, Kendall's tau-c), nominal
by interval(eta), kappa, risk, McNemar, and Cochran’s
and Mantel-Haenszel statistics
v. After all, details are encoded or set, click on ok. Results will appear
in the output window.
MEANS COMMAND
a. It allows you to explore certain characteristics of continuous variables within
those categories. In addition, it allows you to view certain characteristics of
continuous variables by groups.
b. This can be done for various purposes like computing the mean score for
every section, the mean score per gender, etc. It will list the mean, the
standard deviation, and number of subjects for each category.
c. Means procedure also allows the calculation of Analysis of Variance for One-
Way ANOVA and other related statistical procedures
d. An example of this is computing mean total for section by gender. The means
command will show the mean total score , count, and standard deviation of
genders per section, viz:
i. Mean Total Score, N, and SD of Female of Section 1
ii. Mean Total Score, N, and SD of Male of Section 1
iii. Mean Total Score, N, and SD of Female of Section 2
iv. Mean Total Score, N, and SD of Male of Section 2
v. Mean Total Score, N, and SD of Female of Section 3
vi. Mean Total Score, N, and SD of Male of Section 3
vii. Mean Total Score, N, and SD of Female of Section 4
viii. Mean Total Score, N, and SD of Male of Section 4
e. An example of a means command output is presented below:
PROCEDURE FOR MEANS COMMAND:
a. Click on analyze -> Compare Means -> Means
b. In the new window will appear, one can do the following
i. Select the variable he or she wishes to analyze
1. Click on the variable in the left box
2. Click on the arrow to the left of Dependent List
ii. Select the variable for Independent List
1. Click on the variable in the left box
2. Click on the arrow to the left of Independent List
a. Independent list is where you identify the categorical
variables
iii. Select the previous and next to the left and right of layer 1 of 1
1. The arrangement of layers determines the order of layers to which
the variables be arranged
iv. A user can also click on:
1. Cancel to cancel the means procedure
2. Reset to remove all variables and return them to default setting
v. Conduct statistical analysis by clicking “Options”
1. Click on Options
2. In the appearing window, one can click the statistical test to be
conducted
3. One can also determine statistics for first layer
a. Anova table and eta
b. Test for linearly
c. After all variables are set, click on continue then ok.
d. Results of the analysis will appear in the output window
2. Dr. Toob believes that people who meditate regularly are less jealous. She has ten
participants meditate for a week, and ten participants not meditate for a week. After she
measures the level of jealousy for her participants, she uses the means procedure to
determine whether her hypothesis is supported. People who meditated scored 45, 38, 34,
40, 30, 41, 37, 32, 28, and 29. People who did not meditate scored 27, 23, 27, 25, 27,
13, 13, 20, 37, and 23. Write all the SPSS steps commands to perform this analysis. Is
her hypothesis supported? How confident can you be?
a. Opening SPSS
i. Click start in the desktop
ii. Click the IBM SPSS Statistics
iii. Click the IBM SPSS Statist
iv. The initial data screen will appear
b. Defining Variables. In the variable view, define the following variables with the
following descriptions:
Variable Type Width Decimal Label Values Missing
Name
Meditate Numeric 8 0 Did 1 – None
he/she Meditated
meditate? 2 -Did not
meditated
Score Numeric 8 0 None None
c. In the data view, enter the following data
d. Run Means Command. Click on Analyze -> Compare Means - > Means
e. In the appearing window, set variables by clicking on the variable in the left box
then clicking the arrow to where you desire it to be assigned
i. Meditate – Independent List
ii. Score – Dependent List
f. Click on options. In the appearing window, click on Anova Table and eta to conduct
One Way ANOVA on the data. After, click on continue. Then click on ok on the
means window.
g. Results of the analysis will appear in the output window
Those who meditated have a higher mean jealousy score(M=35.40, SD=5.7) than those who
did not meditated(M=23.50, SD=7.106). Dr. Toob hypothesis is that “she believes that
people who meditate regularly are less jealous”. Given that those who meditated have
higher jealousy score mean of 35.40 than those who did not meditated(jealousy mean
score of 23.50), Dr. Toob’s hypothesis is not supported.
However, this conclusion is not confident given no significance level is provided and we
cannot determine if there is a need to reject null hypothesis or not. P=0.001
