Professional Documents
Culture Documents
Schwartz Et Al Normal Eloszlas
Schwartz Et Al Normal Eloszlas
I
n Chapters 3 and 4, we explained the use of SPSS to analyze your research data.
SPSS will do a great job with most data sets, and you will get plenty of output
to share with the world. While you are writing up APA-style results sections and
bragging about your data, do not forget to include the simple parts of the puzzle:
descriptive statistics . When you describe a group of numbers, you can use
values such as measures of central tendency (which tells you one value that
best represents a data set), variability (representing how spread out values in
a data set are) , simple frequency (how often a value occurs in a data set), and
percentages , to name the most popular options. We will cover which options to
use to describe the four types of variables : nominal, ordinal, interval, and ratio.
however, the use of 1 and 2 is arbitrary-we could use any two numbers.
Us ing SPSS
In the next screenshot, we went to Variable View and entered "no tattoo" as
the variable name, and then under Values we entered 1 for no tattoo. Next we
will click Add.
55
56 Y O U R B A S I C S P S S T O O L B O X
14
"
'9
20
�I
2�
n
After clicking the Add button, enter a 2 for tattoo, then c lick Add again
and OK.
10
11
12
13
l'
IS
:7
In Data View, when we View Value Labels, the data would look like the
following.
Te l l Me About It 57
25 : tattoo
1 no t attoo
2 no t attoo
3 no t altoo
t attoo
t aU oo
t aHoo
no t aHoo
8 no t attoo
9 t alloo
10 no t attoo
11 t att oo
12 t attoo
13 t attoo
14 no t attoo
15 t attoo
16 t attoo
17 t attoo
18 t attoo
19 t all oo
20 no t attoo
21 no t altoo
22 t attoo
23 t attoo
24 tatt oo
25 no t attoo
26
27
28
To describe this nominal variable, you might want to look at how often each
value (tattoo or no tattoo) occurs in your sample. You could ask SPSS to provide
a simple frequency for each group by clicking Analyze, Descriptive Statistics, then
Frequencies.
58 Y O U R B A S I C S P S S T O O L B O X
t:<lcq �
110 too
no 1!lOO
110 t 1(00
t,QC
too
100
no tw
to nll t too
t I'QO
� �loo
1:00
no tl1ltoo
I 1too
too
1:Jt!lOO
tllltoo
IIltoo
no tA1too
rnl too
t!lltoo
tlIlroo
umoo
no t�toa
A box will open. Move "tattoo" to the right-side box using the arrow and
then click OK.
la1100
no tonloo
l3tcOO
1 a1loo
Tel l Me About It 59
The output offers a simple frequency for each group as well as percentages.
IORDI:.R=AllALYS l S .
" Frequencies
[t>ataSe';.l]
,,,troo
Valid
I(JSStng
I
I/rtIOO
Cumulil11'o'\J
Pertent
Val,d no laaoo 400
tattoo 1 00 0
Tot.,
With these descriptive statistics, you could begin t o summarize your sample
data. As an additional option, you could offer a measure of central tendency.
Unfortunately, nominal data are so simple that you can only use the mode ,
which is the value that occurs most often in the data set. Although you can easily
see that the mode in our fictional data set is having a tattoo, allow us to show
you how to obtain the information from SPSS.
Because we were already working under Ana/yze, Descriptive Statistics,
Frequencies, return to those options and click Statistics (circled below) on the right
side of the Frequencies box that opens. In the next box that you see, click Mode.
11
"
I)
"
!!t'"
e.,MOtOVClt <$
I� v.." •• grauD �
60 YO U R B A S I C S P S S T O O L B O X
Click Continue and OK to obtain the following output now containing the
mode of 2. Keep in mind that 2 represents the tattoo group, so we know the most
often occurring value is having a tattoo.
/ S'I'AM I ST CS=l-IOD�
/ ORDER=ANALYS I S .
.. Fre q ue n c i e s
( � t.oSe 1 J
t ttao
C umul tr.-e
FleQu ency Percent Valid Perc enl Pe rc ent
altoo 15 60 0 60 0 100 0
Total 25 1 00 .0 1 00.0
The mode, simple frequency, and percentages are the only options that
produce meaningful values with nominal data .
Using SPSS
In Variable View, enter numbers and labels for Values. Notice in the fol
lowing screenshot that 1 logical ly comes before 2 to represent the order of the
categories correctly.
Tel l Me About It 61
Under Data View, you can choose to View Value Labels to see the following.
3S - rank
rank
Jl
1
2--
first -year
--] firsl-year
3 J
-
first -year
4 J fIrst -yea r
5 ---,
-
first-year
() I s o p hom<lle
7 sop homore
8 s op homo l e
9 so p homore
10 so p homore
11 sop homote
12 sop homore
13 junior
14 l Junim
15 Junior
1� Il tni",
62 Y O U R B A S I C S P S S T O O L B O X
Because the data now have a logical order for the categories, you have
more options to describe your data. You can certainly ask for simple frequencies,
and you can ask for the mode. But you can a lso ask for higher-level descriptive
statistics. With ordinal data, the appropriate measure of central tendency is the
median, which is the value at the 50th percentile. If you were to organize values
from lowest to highest (or the reverse order), you could pick out the middle value.
The good news is that SPSS will do the work for you-which is particularly helpful
when you have very large data sets. In addition to a measure of central tendency,
you can report a measure of variability with ordinal data. Simply report the range ,
which in this case for ordinal data will be the lowest and highest values in the
sample. Click Analyze, Descriptive Statistics, Frequencies. Move the "rank" variable
to the right side, click Statistics, and ask for Median, Minimum, and Maximum. A
click in the boxes next to each term will allow you to choose each option.
rU )�..
t ......
\'$" \-
"�'l 1/1
oc�•
•o�
.o�•
•O�.
10 'o�.
aopl""".",
ao�.
1""""
iU"'"
1$ !"tn)t
6 �n�
1""'"
,.....'"
JU"'=I
� !""""
�
22 JIJ"'CI'
23
�4
�
Zti
2]
2$
11......
The minimum and the maximum represent the range on this variable. Yes,
you could ask for the range in SPSS, but remember that we entered values of 1,
2, 3, and 4 for the software to read. If you ask SPSS for the range, it will subtract
1 (the lowest value) from 4 (the highest value) and tell you the range is 3. As
you can see, a range of "3" means nothing. For the range of ordinal data, simply
report the lowest and highest values; SPSS tells you the lowest value is a 1 (which
represents first-year students), and the highest value is a 4 (seniors).
Te l l Me About It 63
• F requ e n c l es
rank
V hrj �'s1''I'eilr S I II 3 ,0 1 11 )
'50;;lhOmOre 7 lO O 10 0 34 ,3
lu la' 10 Jllti 2$ I;. 61 9
SE-II ur 13 J7 1 37 I 1 00 0
J' 35 1 0D O l OO 0
From this output, you can see that the median (middle value) is a 3, which
means the best value to represent this ordinal data set is juniors. And values in
the data set range from first-year students to seniors. To make sense of these data,
you need to remember that 1 first-year students, 2 sophomores, 3 juniors,
= = =
and 4 = seniors.
However, assuming you have interval or ratio data with no outliers, the
appropriate measure of central tendency is the mean (M), and the best measure
of variability is the standard deviation (SD) (although some people use the
standard error of the mean, which is also fine). As an example, imagine you
wanted to know about ratings of happiness among your friends. Let us say that
happiness is operationalized using a rating scale from 1 (Very Unhappy) to 10
(Very Happy), and you asked 17 of your friends to rate their happiness. The data
might look like the happiness ratings below.
7 6 8 7 6 5 4 9 7
8 5 4 6 7 5 8 6
Using SPSS
We can enter the ratings into SPSS. Notice that unlike the nominal and ordi
nal data described above, we have no values to label because the numbers are
not categories; they are simply numerical values.
-�
3 8. 00
7. 00
5 6.00
G 5.00
7 4 .0 0
9.0 0
7 00
8. 00
5 00
4.00
6 00
7. 00
5. 00
8. 00
6 00
18
I
19
Te l l Me About It 65
10
11
12
1)
1.6
11
.e� n' 'I)
:t'R£ C I ES VA.lUABhE.S;:::Happ � � .!
/ STAM
_ _
S � l C S�S7 EV �
I ... �ANALYS1 S •
.. Frequenc ies
� Set 3 ]
Happiness
V 11 17
loAissing o
an
S1d Oenallon
Happiness
V lid 4 00
5 00
6.00
7 00
8.00
9 00
To I 11 H IO O 1 00.0
Notice that because you went through the option of Frequencies in SPSS,
you will get simple frequencies. It is up to you whether you want to examine
that frequency table. Likewise, under Statistics you could ask for Mode, Median,
Range, Minimum, and Maximum (as well as other descriptive statistics options),
but we generally focus on the mean and standard deviation when describing
interval or ratio data with no obvious outliers. In this example of happiness, SPSS
shows us that the average happiness rating from our friends is 6 . 3 5 . (Note: For
APA style, you include only two digits after the decimal point). Recall that happi
ness is measured on a scale from 1 to 10, with higher numbers indicating more
happiness. In addition, the standard deviation tells us that most of our friends rate
their happiness at 6.35, give or take 1 .46 happiness points. We can even add and
subtract the standard deviation of 1 .46 from the mean of 6 . 35 to show a reader
where most of the values fall. That is, most of our friends rate their happiness
between 4.89 (6.35 - 1 .46) and 7 . 8 1 (6.35 + 1 .46) .
Te l l Me About It 67
We should note that SPSS offers a second way to get means and standard
deviations. Click Analyze, Descriptive Statistics, and Descriptives.
!..
2 6 00 1.Ii!ld llOcN
) 8 00
ken
" 00
!lf91 11-fium
� £ 00
li'lllMiIt iJ 2"O PIoII
& 5 00
a8S�
'I
4 00
ytlmll 'iOtl Re<lOOtO<l
9 00
�:!I.
1 00
__amfl1<: Tests
HI 8 00
Fatecu, 0
'I
1 ' 00
11 4 00 �<'JI
iJ 6 00 u_ pC' �Spet'1$&
1
Qu» COtIIfoI
il'd ROC C....�
14 1 00
15 s oo I
16 8 00
11 6 00
Move the variable to the right, click Options, and notice that the default is
for Mean and Standard Deviation (Std. deviation) to be chosen. Minimum and
Maximum are also chosen by default, and we do not mind even though we
usually do not report these values.
68 YO U R B A S I C S P S S T O O L B O X
I Dat.aSet: 3 ]
Of course, the data reported here are the same as under the Frequencies
option, but the output is much more concise. You can use the Descriptives option
for interval or ratio data if you would like.
be concerned that the same people are tested twice; you still have two separate
groups (or samples) of information based on what they were asked to do.
SPSS allows you to click a couple of options after entering your data
(described below), and you will have descriptive statistics for your groups.
Before we go any further, we can look at fictional data. As in our last exam
ple, happiness is operationalized with a rating scale from 1 (Very Unhappy) to
10 (Very Happy) .
Couch Treadmill
3 8
6 7
5 7
7 6
4 9
5 10
8 10
5 8
�
Using SPSS
In this design (the same people in both groups), we will merely enter the
data into two SPSS columns so the data look like what you might write on paper.
If you have forgotten how to label columns, refer to Chapter 6.
, 9 : Couch
Couch Treadmill I var
1 j 3. 00 8 . 00
2 � 6. 00 7. 00
3 5 00 7 . 00
4 7. 00 6. 00
5 � 4 . 00 9. 00
6 5 00 1 0 . 00
7 8 00 1 0 .00
8 5 00 8 00
9
n
70 YO U R B A S I C S P S S T O O L B O X
To ask SPSS for descriptive statistics on your two groups, click Analyze,
Descriptive Statistics, Descriptives as we covered in the prior section.
In the box that opens, move the two variables over to the right under
Variable(s) by highlighting them and clicking the arrow in the middle of the box.
Then click Options to make sure Mean and Standard Deviation (Std. deviation)
are checked.
3 00 8 00
6 00 7 00
5 00 7 00
1 00
' 00
S OO
8 00
5 00
Te l l Me About It 71.
[ DataSet O ]
N Minimum
8 3.00
8 6.00
S
Notice that we now know the average happiness rating in each experimental
condition. We also have a number (the standard deviation) for about how spread
out those values are within a condition. When people sat on a couch, the average
happiness rating was 5 . 38, give or take 1 .60 happiness points on a scale from 1
to 1 0 , with higher number indicating more happiness. On the other hand, when
they worked out on a treadmill, the average happiness rating was 8 . 1 3, give or
take 1 .46 happiness rating points .
S u m mary
The descriptive statistics you choose rely o n your variables' levels o f measure
ment: nominal, ordinal, interval, or ratio. Be sure you know the level of your data
before summarizing it, then describe your sample or samples with appropriate
descriptive statistics.
In this book, you will learn how to write APA-style results sections for many
statistical analyses, but keep in mind that you should always run descriptive
statistics to describe your variables to the reader. Those descriptive statistics are
included in the results section in some way. Here are some options:
For all of these options, rely on SPSS output to give you the data you need
to report. If you choose to create an APA-style table, refer to the APA manual
for instructions or take a look at An EasyGuide to APA Style (Schwartz, Landrum,
& Gurung, 20 1 4) . If you choose to create a figure such as a graph, see the APA
manual as well as An EasyGuide to Research Presentations (Wilson & Schwartz,
20 1 5) . Whenever possible, offer your reader a table or figure for important
results. In particular, figures provide information in an interesting format and
give readers a break from reading text.