Next: Testing for normality, Previous: Inverting negatively coded variables, Up: Data Screening and Transformation [Contents][Index]
A sensible check to perform on survey data is the calculation of
reliability.
This gives the statistician some confidence that the questionnaires have been
completed thoughtfully.
If you examine the labels of variables v1, v3 and v4,
you will notice that they ask very similar questions.
One would therefore expect the values of these variables (after recoding)
to closely follow one another, and we can test that with the RELIABILITY
command (see RELIABILITY).
Example 5.4 shows a PSPP session where the user (after recoding
negatively scaled variables) requests reliability statistics for
v1, v3 and v4.
PSPP> get file='/usr/local/share/pspp/examples/hotel.sav'. PSPP> display dictionary. PSPP> * recode negatively worded questions. PSPP> compute v3 = 6 - v3. PSPP> compute v5 = 6 - v5. PSPP> reliability v1, v3, v4. Output (dictionary information omitted for clarity): Scale: ANY Case Processing Summary +--------+--+-------+ |Cases | N|Percent| +--------+--+-------+ |Valid |17| 100.0%| |Excluded| 0| .0%| |Total |17| 100.0%| +--------+--+-------+ Reliability Statistics +----------------+----------+ |Cronbach's Alpha|N of Items| +----------------+----------+ | .81| 3| +----------------+----------+ |
Example 5.4: Recoding negatively scaled variables, and testing for
reliability with the RELIABILITY
command. The Cronbach Alpha
coefficient suggests a high degree of reliability among variables
v1, v3 and v4.
As a rule of thumb, many statisticians consider a value of Cronbach’s Alpha of 0.7 or higher to indicate reliable data. Here, the value is 0.81 so the data and the recoding that we performed are vindicated.