the sign test and chi squared

Created by

jenelle

Cards (7)

we use the sign test
when a test for difference between conditions has been carried out
when nominal data has been gathered
when the experiment used a repeated measures/matched pairs (related design)
you have to work out the calculated value for sign test (calculated s value)
calculate difference between 2 conditions (condition A minus condition B)
record difference
for each calculated difference, determine whether the difference is positive, negative or zero
count total number of positive differences and total number of negative differences (ignore zero differences)
calculated value of s is the smallest of the two counts
to interpret the sign test
calculated value
hypothesis (one or two tailed)
experimental design
critical values table (given to you)
level of significance (if not given, assume it's 0.05)
no. of participants (found on stem)
model answer
3 (calculated value s) > 1 (critical value s)
results are not significant as the calculated value of s is higher than the critical value of s
null hypothesis has to be accepted
model answer - explain why sign test was chosen
the researcher used nominal data by classifying patients into categories like 'improved', 'deteriorated' or 'neither'. secondly the study employed a repeated measures design (related), meaning the same patients' records are compared before and after treatment
model answer - is the data significant
as the calculated value (5) is higher than the critical value (4), the result is not significant at the 0.05 level meaning we have to accept the fact that CBT may not be effective as a treatment to depression.
to interpret chi-squared
calculated value x^2 (observed value) - will always be given to you
hypothesis (one or two tailed)
experimental design
critical values table
level of significance (if not given assume it's 0.05)
table of results known as a contingency table
formula to calculate df - it will also be given