Research Methods in Management JRMR25 – 2022

1
A3: Quantitative analysis

General info:
- Don’t be alarmed. Once you are in front of SPSS, the analyses will be fairly swift.
- Share the responsibility. This assignment is partly technical, but also about the interpretation
of the results: how are the statistical results “translated” into some meaningful statements
about the situation in the “real” world that produced the numbers.
In this assignment we will work with a data file called “DATA22.SAV” and look into
explanations for salary differences. In particular we will examine the issue of gender differences.
We can sum this up with the following research questions:
1) What effect does gender have on the current salary of the employees?
2) How large are these differences?
The data contains descriptive information about 474 employees of an actual company. While the
data are real, they are also old. So, do not draw any conclusions about today’s situation from this
data set. At the time though, they were very valid. Food for thought…
How to submit your answers
Export the results, add text comments and submit as pdf. SPSS can export fairly well to MS Word so I
suggest you do that and export to pdf.
Please keep in mind that SPSS “remembers” what variables you used in your last analysis.
Therefore, clear any selected variable before doing new analysis unless you what do duplicate
the last analysis. Otherwise, stranger things may happen.
Open the file “DATA22.SAV”.
“Salnow” is measured in US dollars per year (Remember that the data file isn’t exactly new!).
“Edlevel” is measured in number of years of formal education.
Questions
1. Which is the dependent variable in this analysis and what type is it (ratio, interval, ordinal or
categorical)? (1 p)
2. The coding of the categorical variable “gender” is of a particular type, using numerical 0/1 to
code the categories. What is this type of variable formally called? (1 p)
3. Is the distribution of variable “Current salary [salnow]” skewed to the left (negative), to the
right (positive) or symmetrical? Motivate your answer and copy the tables with the measures
of central tendency into your text file. [Tip: Export results from SPSS as “RTF”. They tend to come
out nicer that way.] (2 p)
[Hint: Do MENU Analyze=>Descriptive statistics => Frequencies and select the appropriate
measures of central tendencies. Also click “Charts” and select “Histograms”.]
4. What is the difference between the mean salaries of men and women in this firm? (1 p)
Research Methods in Management JRMR25 – 2022

2
[Hint: Choose the correct test for testing differences in means between two groups in a
sample. The test variable is “Current salary” and the grouping variable is “gender”.]
5. Do a linear regression with the “Current salary [salnow]” and “Gender of employee
[gender]”. Look at the output. Interpret the constant and the coefficient in the box labeled
“Coefficients”. (You need to understand the difference between a dependent and
independent variable.) (1 p)
[Hint: Do MENU Analyze=>Regression => Linear. Add the variable “Current salary
[salnow]” and “gender” into the appropriate boxes. For this you need to decide which is the
dependent variable which is the independent variable. For interpreting the coefficient, you may look
at the output from the statistical test in Q4 and also check in the SPSS “Data view” how the
variable “Gender of employee [gender]” is coded.
6. Gender is probably not the only thing that matters for your salary. So, we will estimate the
following model:
a) Run the test for this model using the information from Q5. Does this model do a worse
or better job in explaining wage differences than the model we estimated for Q5? Hint:
This is not about statistical significance but rather about “goodness-of-fit”. (1 p)
b) What is the implicit null hypothesis for the regression coefficients? (1 p)
c) Compare the results to the analysis you did for Q5. The coefficient for “Gender of
employee” has decreased. Why? To answer this question, you may want to check if “level
of education” and “gender” are correlated. Provide an answer based on the statistical
analysis and interpret the results. Be clear and through in your answer. (2 p)
Good luck!
Current salary
Gender
Level of
education