SOC4217 QUANTITATIVE RESEARCH METHODS
Computer Workshop:
Exploring Data With Graphs
You should work through the instructions for this workshop systematically. You are also
encouraged to familiarise yourself with the SPSS interactive “help” system. Context-sensitive help
is available as an option in almost all dialogue boxes. In addition, the general “Help” menu
contains step-by-step tutorials and case studies on the use of vital statistical techniques - including
all the techniques covered in this module - as well as a comprehensive reference to specific SPSS
topics and command syntax.
SYMBOLS USED IN THIS DOCUMENT:
General instruction
Instruction for button clicking
Offers additional explanations to you
A double outlined box indicates the window you should see at this point
A single outlined box indicates the dialogue box you should see at this point
In this workbook, you will perform fundamental descriptive analysis using standard measures of
central tendency and dispersion.
KEY LEARNING OUTCOMES OF THIS WORKSHOP:
1
Bar charts – histograms – box plots in SPSS
Part A. Downloading the data
What is the general state of health of respondents in the PSEHK 2012 data?
Assign the scale of measurement and missing values as appropriate.
Get the frequency table.
How did experiencing pain interfere with respondents’ normal work (including work outside
the home and housework)?
Assign the scale of measurement and missing values as appropriate.
Get the frequency table.
Part B. Producing graphs with the Frequencies command
You can draw bar charts through the Charts option in the SPSS frequencies procedure.
Go to Analyze | Descriptive Statistics | Frequencies
Transfer appropriate variables to the Variable(s) box.
Click the Charts button on the right-hand side.
Select Bar Charts and leave the default Chart Values option of Frequencies
2
Plotting Percentages might also be helpful in some instances – especially if you are
thinking to present your bar chart in a research report.)
Click Continue.
Click OK.
Find out the mean, median, maximum, minimum and the standard deviation of how much
HK$ per month (after tax) respondents think is necessary to keep a household out of poverty.
This is a continuous variable, and you should switch off the Display Frequencies Tables
option in the Frequencies Dialogue Box (if you are not sure why just leave it and see what
happens).
Once you are happy with your descriptive statistics, draw a histogram by deselecting Bar
Charts and select Histograms within the Charts options.
What does the histogram tell you about the skewness and kurtosis of this variable?
Once you are happy with your answer (you should check with your tutor if you are not sure),
you should go ahead and find one more continuous variable in the PSEHK2012 data. Then,
repeat the steps from above and compare its distribution against the variable from above.
By ticking the box within the Charts options, you can also tell SPSS to ‘show a normal
curve on the histogram’ – this might make answering the above question a little easier for
you.
Part C. Drawing a box plot
You already learned today how to draw bar charts and histograms with SPSS. Drawing box plots is
also not complicated, but unfortunately, this cannot be done directly through the Frequencies or
Explore commands. Instead, it is best to use the SPSS Graph command. To be more precise, we
will be using a function called Legacy Dialogs.
3
Go to Graphs | Legacy Dialogs | Boxplot (as shown below)
You should now choose the Simple option and select Summaries of separate variables.
Click Define to get to the next step.
SPSS gives you a vast range of options to draw your box plots, but the simplest boxplot will suffice
for our purpose. All you have to do now is to select your variable and move it into the ‘Boxes
Represent’ box and
Click OK.
Your box plot should look like the one below. Take a moment now to try and interpret it. What does
the box plot tell you about the distribution of values in this variable? Are there any extreme cases or
outliers in your data? Can you identify the median and interquartile range?
4
Part D. Practice Exercise
There is only one exercise today, and we will talk about this again at the beginning of next week’s
lecture.
You have learned many different descriptive techniques over the last two weeks, but how would
you write up your descriptive results in a concise and informative manner?
Using one categorical and one continuous variable in the PSEHK 2012 dataset of your choice, you
should try and do this on no more one A4 page (excluding the raw SPSS output) to practice
your data interpretation and writing skills. Be extremely careful when assigning scales of
measurements and missing values, and please present all tables and charts in an Appendix to your
write-up.
*********************END*********************