47-512 Applied Research Methods
due in class 2/20/03
Purpose.
You may want to try running
the SPSS tutorial the first time you use the program.
How to start. Open SPSS
and select type in data. You
will then have a screen facing you that will permit you to type in any kind of
data you choose (data view). If
you click down on the tabs on the lower left of the screen, you can switch
between screens that permit you to store information about your variables (variable view) and to make
adjustments (i.e., change the number of decimal places showing from 2 to 0 when
you don’t need them as in this exercise).
Variable names need to have 8 characters or less; longer and more
descriptive titles can be used as labels. These are assigned by typing into the appropriate spaces on the
variable view screen.
Remember: Each
participant gets one row, and one row only.
Each variable gets one column.
After you have started
typing, save early and save often. Note
that data sets are saved as xxx.sav files.
Outputs are saved as xxx.spo.
These will be the primary kinds of files you will use. [Note that under FILE, SPSS is also
capable of opening data that are stored in Excel or other spreadsheet forms or
in text. Once you open these data sets
as an SPSS data file, you may use SPSS to explore the data.]
After the data are entered,
check them. Then double check
them. Any results you get from data
entered in error are going to be in error as well.
In this exercise, you are
being asked simply to summarize the results of a survey or 25 teachers, by
providing the most informative descriptive statistics about each of three
variables. You want to treat these
individuals as a sample representative of other people like them.
What to do? The first
decision you must make is the type of data that you have. If it is nominal or categorical, you will
use proportions (percentages) to describe them. If they are ordinal or ratio, you will use measures of central
tendency and spread. You must then choose
between reporting (a) means and standard deviations, or (b) medians and
interquartile ranges.
How to decide. Before
computing a number, look at a picture.
Look at the distribution of years of education and class size by going
to GRAPHS and choosing histogram. For another view, go to GRAPHS, choose Boxplot and
create a simple boxplot for separate variables.
Forgotten what these
plots tell you? Go to HELP and open the topic. Opening
boxplot, for example, allows you to be reminded that, “Boxplots show the
median, interquartile range, outliers, and extreme cases of individual
variables.”
What are you looking for? You use the
mean and standard deviation if the distribution approximates a normal one—the
histogram should look like a bell shaped curve and the boxplot should be
symmetrical without extreme data points (outsiders). If the histogram is very uneven or shifted to one side or another
(skewed) or if there is marked asymmetry or are substantial outliers on the box
plot, you won’t get very good information from the mean and standard deviation,
so use the median and interquartile range.
To find proportions of
categorical data, select Analyze then Descriptive Statistics then
Frequencies. The frequency table
will list each represented value of a variable and indicate what proportion of
the total have that value (e.g., green .90, red .02, yellow .08) along with the
cumulative proportion for the values when they are sorted in order
(alphanumeric).
To find the median and
interquartile (25th and 75th percentiles) and the mean
and standard deviation, again select Analyze then Descriptive
Statistics then Frequencies.
Move the variable name into the box of variables to be analyzed and then
choose Statistics and deselect the frequency table. The Statistics screen will allow you to
select the particular descriptive statistics you want to compute: mean, median, standard deviation, and
quartiles.
For this assignment, select the appropriate descriptive statistics for each of the variables listed in the following table. It is not imperative that you use SPSS for this exercise; you may complete it by hand if you would like. Either way, please show all your work (print the data screen and your output). Also, on one page write down the (a) name of the variable, (b) the descriptive statistics, and (c) the reason you chose that particular way to describe that particular variable.
The data (absolutely hypothetical): Given the current budget crisis, 25 teachers were asked what they would recommend the Massachusetts Teacher’s Union advocate as a target enrollment for elementary school classrooms that would balance financial constraints with educational concerns. They were also asked their own educational background and their political party affiliation (democrat, republican, or independent).
|
Subject |
Political affiliation |
Years of Education |
Class Size |
|
101 |
D |
16 |
21 |
|
102 |
D |
16 |
24 |
|
103 |
D |
16 |
21 |
|
104 |
D |
21 |
20 |
|
105 |
D |
16 |
21 |
|
106 |
R |
16 |
22 |
|
107 |
R |
23 |
21 |
|
108 |
I |
22 |
24 |
|
109 |
R |
16 |
15 |
|
110 |
D |
16 |
20 |
|
112 |
I |
21 |
20 |
|
111 |
R |
17 |
18 |
|
113 |
D |
17 |
22 |
|
114 |
D |
16 |
16 |
|
115 |
R |
16 |
18 |
|
116 |
R |
17 |
22 |
|
117 |
D |
28 |
24 |
|
118 |
D |
16 |
26 |
|
119 |
R |
19 |
20 |
|
120 |
R |
17 |
21 |
|
121 |
I |
16 |
20 |
|
122 |
R |
17 |
18 |
|
123 |
D |
24 |
22 |
|
124 |
D |
17 |
21 |
|
125 |
R |
16 |
22 |