Practice Data

Internet usage for personal (nonprofessional) reasons

An internet service provider commissioned a study to investigate the determinants of internet

usage for personal (nonprofessional) reasons. A sample of 100 respondents is drawn randomly
from the population. For illustrative purposes, we consider only 30 observations. Table 1
contains the data for 30 respondents giving the gender (1 = male, 2 = female), familiarity with
the internet (1 = very unfamiliar, 2 = unfamiliar, 3 = slightly unfamiliar,……., 7 = very familiar),
internet usage for personal reasons in hours per week, attitude toward the internet and toward
technology, both measured on a 7-point scale (1 = very unfavorable, 2 = unfavorable, 3 = slightly
unfavorable, …….., 7 = very favorable), whether the respondents have done shopping on the
Internet (1 = yes, 2 = no), and whether the respondents have done banking on the Internet (1 =
yes, 2 = no).

Variables : 8 Cases: 30

Variable Information:
Name Position
RES_NUM (nominal) *No label* 1
IUPW Internet usage in hours per week 2
ATI Attitude towards internet 3
Value Label
1 very unfavorable
2 unfavorable
3 slightly unfavorable
4 neither unfavorable nor favorable
5 slightly favorable
6 favorable
7 very favorable
ATT Attitude towards technology 4
Value Label
1 very unfavorable
2 unfavorable
3 slightly unfavorable
4 neither unfavorable nor favorable
5 slightly favorable
6 favorable
7 very favorable
FWI Familiarity with the internet 5
Value Label
1 very unfamiliar
2 unfamiliar
3 slightly unfamiliar
4 neither unfamiliar nor familiar
5 slightly familiar
6 familiar
 7 very familiar
G Gender 6
Value Label
1 Male
2 Female

SOI Whether done shopping on internet 7

Value Label
1 yes
2 no
BOI Whether done banking on internet 8
Value Label
1 yes
2 no

Exercises

1. Define the variables and complete the data entry.

Done
2. Find the minimum and maximum value for each variable in the study.
It makes sense for the scale variables, but for now we are applying to all the
variables.

Familiarity can’t be 9 (10th respondent) . So, this is the issue we need to identify
from this step. The data entry person or the respondent made the mistake.

N in ATI (17th respondent) and Gender is 29 (30th respondent), which means one
response is missing. And Valid N is 28, that means we have complete data of 28
respondents only.
 For small data, we can identify the problem with which respondent. For large data
sets, you can use transform, create another variable. Identify the issue and modify
the response by cross checking. Later clear the variable from variable view.

3. Fix the missing values in the data.

Data Cleaning or data pruning. We have to go to analyze, descriptive statistics and
frequencies and find the missing values

Go to Transform -> Compute variable -> Function group – All – Missing(ATI)

Click Ok. And check which response has the 1 for transform variable created. That
is the one missing response and cross check and give the value.

Else, replace with the mean value. For scale data, we can do this.
 For nominal we can’t do mean. We should cross check. Assume that you go back to
questionnaire and find the missing value and fill that.

4. Construct a frequency distribution of the variables Gender, Familiarity with the internet,
whether done shopping on internet and whether done banking on internet.

Analyze -> descriptive statistics -> frequencies. First give familiarity with the
internet variable and use chart -> histogram (for scale data we use histogram). Click
continue and then ok to get the output. Output is saved.

Later do it for other 3 variables combinedly selecting pie chart. Refer output.

You can also edit the chart by double clicking.

5. Make a pie chart of the Gender, Usage of Internet: Shopping and Usage of internet:
Banking.

Refer output

6. Construct a frequency distribution of the variables Internet usage, Attitude Toward

Internet and Attitude Toward Technology.

Refer output
7. Use explore feature to make a box and whisker plot and stem and leaf plot for the
variable Internet usage.

Go to plots and select box plot frst option and stem & leaf, click continue and ok.
Refer the output sheet.
 Similarily you can also see for different segments based on gender.

Click on plots, select box and stem and leaf and click continue and then ok. Refer
the output sheet.

8. Compute the descriptive statistics for Internet usage, Attitude Toward Internet and
Attitude Toward Technology.

Refer output sheet.

9. Do a cross tabulation between Gender and Internet usage.

Cross matrix between two categorical variables, to see the association between them.

We tried between Gender and whether done shopping online.

Column – independent variable

Later tried to get only percentage and not counts. Refer the output for both plots.
Now, gender and internet usage.

But internet usage is scale variable. So, we have to convert that to categorical
variable by recoding it as high usage if more than 5 and low usage otherwise.

Go to transform -> recode into different variables.

Continue and Ok. New variable will be created with 2 and 1. Where 1 represents
less usage and 2 represents high usage. Save it. Now go to cross tabs and plot.
 Refer the output sheet.

10. Do a multiple cross tab between Gender, Familiarity with the Internet and Internet usage

To see if gender is influencing the internet usage or the familiarity with the internet.
Change familiarity with internet as the categorical variable. Transform it as low
familiarity if less than or equal to 4 and more familiar if more than 4.01.
Then do the cross tabs by adding the layer as Familiarity with internet categorical
variable. Refer the output sheet.
11. Do a chi-square test between Gender and Internet usage.

Hypothesis Testing – Association based (Chi-Test)

- Mean based (T-test)

Less than 0.05, then reject the null hypothesis. Here we should refer to Asymp
significance – Continuity correction value as p value. In the below screenshot and refer
output sheet.
12. Do a chi-square test between Familiarity and Internet usage.

Do same as above question and use familiarity with internet in place of gender.
Respondent Internet Attitude Attitude Usage of Usage of
Number Usage Toward Toward Internet: Internet:
Internet Technology Familiarity Gender Shopping Banking
1 14 7 6 7 1 1 1
2 2 3 3 2 2 2 2
3 3 4 3 3 2 1 2
4 3 7 5 3 2 1 2
5 13 7 7 7 1 1 1
6 6 5 4 4 2 1 2
7 2 4 5 2 2 2 2
8 6 5 4 3 2 2 2
9 6 6 4 3 2 1 2
10 15 7 6 9 1 1 2
11 3 4 3 4 2 2 2
12 4 6 4 5 2 2 2
13 9 6 5 6 1 2 1
14 8 3 2 6 1 2 2
15 5 5 4 6 1 1 2
16 3 4 3 4 2 2 2
17 9 3 6 1 1 1
18 4 5 4 4 1 1 2
19 14 6 6 7 1 1 1
20 6 6 4 6 2 2 2
21 9 4 2 6 1 2 2
22 5 5 4 5 1 2 1
23 2 4 2 3 2 2 2
24 15 6 6 7 1 1 1
25 6 5 3 6 2 1 2
26 13 6 6 6 1 1 1
27 4 5 5 5 2 1 1
28 2 3 2 4 2 2 2
29 4 5 3 4 1 1 2
30 3 7 5 3 1 2