Statistical tests - A-level Flashcards
when is Mann Whitney U used
-unrelated data (independent groups)
- ordinal data
-test of difference
how is the data used in Mann Whitney U
in a table of ranks
what is a table of ranks in Mann Whitney U
lowest is given a rank of 1. if multiple data points are the same value the positions they would be are added and mean is calculated
how can ranks in Mann Whitney U be shown
frequency table
what is the value of u in Mann Whitney U
calculated and is the smaller of the values
how is Mann Whitney U seen as being significant
if the value of u is less than the critical value so the null hypothesis can be rejected and alternative hypothesis accepted
when is Wilcoxon used
-looking for a difference
-related design (repeated measures)
-ordinal data
how is ranking used in Wilcoxon
-done between difference in the two sets of data (signs are ignored during ranking)
-if the difference is 0 the data is not included in the ranking and is deducted from the N value
what is the T value in Wilcoxon
calculated value
how is the T value calculated in Wilcoxon
sum of the less frequent sign
how is Wilcoxon seen as significant
is the calculated value of T is less than the critical value the null hypothesis can be rejected and the alternative hypothesis accepted
is Mann Whitney U and Wilcoxon parametric tests
No, they are non-parametric tests
when is an unrelated t-test used
-looking for a difference
-interval data
-unrelated data (independent groups)
-criteria for parametric test must also be filled
what is the criteria for a parametric test
-level of measurement is interval
-normal distribution
-homogeneity of variance (as standard deviations of both groups are similar)
table of data in an unrelated t-test
-calculate sum of scores for A
-repeat for B
-square each value in the A group and calculate the sum of the squared values
-repeat for group B
what is the calculated value in an unrelated t-test
t value
when is an unrelated t-test significant
if the calculated value of t is greater than the critical value the difference is significant and the null hypothesis can be rejected and alternative hypothesis can be accepted
are the related t-test and unrelated t-test parametric tests
yes
criteria for a related t-test
-looking for a difference
-related design (repeated measures)
-interval data
-requirements for a parametric test must also be met
how is the data table used for related t-test
-calculate the difference between scores for condition A and B
-add the sum of the differences
-square each difference
-add the sum of the squared differences
what is the calculated value called in a related t-test
t value
when is results from a related t-test significant
when calculated value of t is greater than the critical value the difference is significant so the null hypothesis can be rejected and the alternative hypothesis accepted
are pearsons and spearman’s tests of correlation
yes
what is the criteria for spearman’s rho
-looking for a correlation
-ordinal data in one or bother is the variables) can be used with interval data)