Reliability Analysis
The reliability of a
measuring instrument is
defined as the ability of
the instrument to measure
consistently the
phenomenon it is
designed to assess.
Reliability, therefore,
refers to test consistency.
Cronbach’s Alpha
It is very common in research to collect
multiple measures of the same construct.
For example, in a questionnaire designed
to measure optimism, there are typically
many items that collectively measure the
construct of optimism. To have
confidence in a measure such as this, we
need to test its reliability, the degree to
which it is error-free. The type of
reliability we'll be examining here is
called internal consistency reliability, the
degree to which multiple measures of the
same thing agree with one another.
3
Cronbach’s alpha—This is a single
correlation coefficient that is an
estimate of the average of all the
correlation coefficients of the items
within a test. If alpha is high (0.80 or
higher), then this suggests that all of
the items are reliable and the entire
test is internally consistent. If alpha
is low, then at least one of the items
is unreliable, and must be identified
via item analysis procedure.
Cronbach’s Alpha
Example: Does my questionnaire measure customer
satisfaction in a useful way? Using reliability analysis,
you can determine the extent to which the items in
your questionnaire are related to each other, you can
get an overall index of the repeatability or internal
consistency of the scale as a whole, and you can
identify problem items that should be excluded from
the scale.
5
What is alpha and why should we care?
– Cronbach’s alpha is the most commonly used
measure of reliability (i.e., internal
consistency).
– It was originally derived by Kuder &
Richardson (1937) for dichotomously scored
data (0 or 1) and later generalized by
Cronbach (1951) to account for any scoring
method.
– A high value of alpha is good, but it is
important to have a deeper knowledge to use
it properly.
Other types of reliability
– Test/Re-Test
» The same test is taken twice.
– Equivalent Forms
» Different tests covering the same topics
» Can be accomplished by splitting a test into
halves
Cronbach’s basic equation for alpha
n Vi
1
n 1 Vtest
– n = number of questions
– Vi = variance of scores on each
question
– Vtest = total variance of overall scores
on the entire test
How alpha works
– Vtest is the most important part of
alpha
n Vi
1
n 1 Vtest
– If Vtest is large, it can be seen that
alpha will be large also:
» Large Vtest Small Ratio ΣVi/Vtest
Subtract this small ratio from 1 high alpha
What makes a question “Good” or “Bad” in
terms of alpha?
– SPSS will report “alpha if item deleted”,
which shows how alpha would change if that
one question was not on the test.
– Higher “alpha if item deleted” means a
question is not so good because deleting
that question would improve the overall
alpha.
Obtaining Reliability of
Scale - Example
105 members of the community
completed a ten-item ‘attitudes-to-
help-seeking’ instrument, using a 5-
point Likert scale (1=Strongly disagree
to 5=Strongly agree). You wish to
determine the internal consistency of
this scale using Cronbach’s alpha.
The data is given in
Attitude_Reliability.sav
Obtaining Reliability of
Scale using SPSS -
Example
Obtaining Cronbach’s Alpha:
Analyze → Scale →Reliability Analysis
Model → Alpha
Drag items on the left to the right
dialog box
– In this case: items hs1 to hs10
Reliability
Press “statistics”
Choose “scale if item deleted”
Then Press “continue”
Then “OK”
Reliability
The SPSS Output
Case Processing Summary
N %
Cases Valid 105 100.0
Excludeda 0 0.
Total 105 100.0
.a. Listwise deletion based on all variables in the procedure
105 cases (observations/respondents) were used
in the calculation of Cronbach’s alpha
Reliability Statistics
Cronbach's Alpha
Based on
Standardized
Cronbach's Alpha Items N of Items
768. 792. 10
The obtained alpha score is 0.768, which indicates that the
scale has high internal consistency (reliability).
Cronbach alpha is a reliability coefficient that
indicates how well the items are positively
correlated to one another. The closer the
Cronbach alpha is to 1, the higher the internal
consistency. As a thumb rule, values of below
0.6 are considered to be poor, 0.6 to 0.7
ranges are acceptable and those over 0.7 are
good.
Item-Total Statistics
Scale Mean Scale Corrected Squared Cronbach's
if Item Variance if Item-Total Multiple Alpha if
Deleted Item Deleted Correlation Correlation Item Deleted
hs1 17.94 21.862 648. 528. 720.
hs2 17.39 23.048 345. 301. 764.
hs3 18.19 22.348 693. 533. 719.
hs4 17.72 22.836 379. 259. 758.
hs5 17.75 21.977 646. 464. 721.
hs6 17.86 23.220 425. 209. 749.
hs7 17.39 25.029 246. 158. 772.
hs8 18.45 24.692 415. 339. 752.
hs9 18.15 24.265 526. 389. 743.
hs10 17.44 23.710 253. 193. 781.
Note that items 7 and 10 have lowest “corrected item –
total correlations”. If these two items were removed from
the scale, the “alpha if item deleted” column shows that
overall reliability would increase slightly. Hence, deletion
of these items may be considered.