Calculations and tests Flashcards
(14 cards)
How do you calculate variance
1) Calculate the mean score for each condition
2) Subtract the mean score from each of the individual scores
3) Square this value
4) Add them up
5) Divide by number of participants present
How do you calculate SD
1) Calculate mean score for each condition
2) Subtract mean score from each given score
3) Square this value
4) Add them up
5) Divide by number of participants
6) Take the root of this value
What does a Spearman rho test find?
Correlation coefficient
When should a spearman rho test be done
The study has ordinal data
The study is correlational
Repeated measures design
Spearman rho procedure
1) Place data from each participant into a table
2) Rank the data for each variable separately. Give the lowest value a 1. When scores share a value give them the middle rank
3) Find the difference between rank A and rank B (B-A)
4) Square the difference
5) Calculate the R value with the formula
R= 1- (6Sumd^2)/n(n^2-1)
6) Compare value to table of values
When would you use a Wilcoxon signed rank test?
Experiments
Repeated measures or matched participants
Ordinal level data has been collected
Wilcoxon procedure
1) Remove all scores which have the same in each condition
2) Calculate the difference in scores between condition A and condition B (A-B)
3) Rank the differences, ignoring the negative signs
4) Add together the value of the rankings of the LEAST common sign to get the T value
5) Compare the T value to the T value in the table in the booklet. Assume 2 tailed hypothesis if not mentioned
If the T value is less than or equal to the critical value, the results are significant
When is a Mann Whitney U test used?
Independent measures
Ordinal data
Mann Whitney procedure
1) Place both groups in order from largest to smallest
2) Order the scores in a SINGLE list (Coloured pens)
3) Rank those new ordered scores
4) Calculate a āUā value for each group. This is done by taking the sum of the ranks for the group and subtracting n(n+1)/2
5) Choose whichever U value is the lowest. This is the calculated U value.
6) Compare this to the critical value in the table. If U is Less than or Equal to the critical value then U is significant
What is a parametric test
Have interval or ratio level data
Have a normal distribution of data
When is a Binomial sign test used
Nominal Data
Repeated measures
Procedure of a binomial sign test
1) Calculate the difference in the scores of the conditions (B-A)
2) Add up the number of positive and negative signs from these differences. The smaller number is the observed value
3) Compare this observed value to the critical values in the book.
If the observed value is Smaller or Equal to the critical value it is significant
When would you use the Chi Square test?
Nominal data
Independent measures
Chi Square procedure
1) Add up the scores for each row and column
2) Calculate the expected frequency for each (Row total x Column total)/Overall total
3) Find the difference between observed and expected frequencies
4) Square this difference
5) Divide the squared differences by the expected frequency for each cell
6) Add up all the values to get the calculated Chi value
7) Find the degrees of freedom (rows-1)x(columns-1)
Compare to table. If calculated value is HIGHER then it is significant
