Statistical Tests Flashcards
Mann Whitney U Test
Criteria
- Ordinal/ Interval
- Independent Measures
- Test of Difference
Mann Whitney U Test
Procedure
- Rank all scores from both groups
- Put the ranks back into the original columns and count up total rank for each column (condition)
- Find observed value (U) by using the formula provided with the data from the column with the lowest total value
- Find critical value using number of participants and 0.05 significance.
Mann Whitney U Test
Significance
Observed Value must be less than or equal to Critical Value to accept the experimental hypothesis
Significant
Wilcoxon Signed Rank Test
Criteria
- Ordinal/ Interval
- Repeated Measures
- Test of Difference
Wilcoxon Signed Rank Test
Procedure
- Find difference between between two scores, could be positive or negative
- Rank the differences ignoring the signs
- Count how many positives and negatives
- Add up to ranks for the least common sign for the observed value
- Number of participants who had a difference is the value for N
Wilcoxon Signed Rank Test
Significance
Observed Value must be less than or equal to Critical Value to accept the experimental hypothesis
Significant
Chi Squared Test
Criteria
- Nominal Data
- Independent Measures
- Test of Difference
Chie Squared Test
Procedure
- Create total for all columns and rows
- Calculate expected frequency by multiplying row total by column total and dividing by overall total (for specific cells)
- Use the formula provided to calculate the observed value
- Calculate degrees of freedom by multiplying number of rows- 1, by number of columns- 1
- Find critical value with degrees of freedom and 0.05
Chi Squared Test
Significance
Observed value must be larger or equal to critical value to accept the experimental hypothesis
Significant
Binomial Sign Test
Criteria
- Nominal Data
- Repeated Measures
- Test of Difference
Binomial Sign Test
Procedure
- Add signs to the data columns, + if yes and - if no, ignore if there’s no change
- Count up how many of each sign
- The smallest number of the two becomes your observed value
- Find critical value using the number of participants were there was a difference
Binomial Sign Test
Significance
Observed Value must be less than or equal to the critical value to accept the experimental hypothesis
Significant
Spearmans Rho
Criteria
- Ordinal Data
- Correlation Design
- Exploring relationships between two co variables
Spearmans Rho
Procedure
- Individually rank the two separate columns, will end up with four instead.
- Find d by calculating the difference between the ranks
- Find d2 by squaring each difference
- Find the total of all the d2 by adding them together
- Use formula provided, n is the number of participants
- A strong positive correlation is close to +1 and a strong negative is close to -1
- Find critical value using n and 0.05
Spearmans Rho
Significance
Observed value must be higher or equal to the critical value in order to accept the experimental hypothesis
Significant
Type 1 Error
False Positive
Accept the experimental hypothesis when it was wrong
Type 2 Error
False Negative
Accept the null hypothesis when it was wrong
Column and Row order for the statistical test names
RIC
NOI
Statistical Test Table Sentence
Big Cats Can Wear Multiple Sunglasses … … Period
Parametric Assumptions
- Populations should be normally distributed
- Variances of populations should be approx. equal
- Should have at least interval or ratio data
- No extreme scores
Nominal Data
Named data which can be separated into discrete categories that do not overlap, i.e. Male or female
Ordinal Data
Data that can be placed into some kind of order, but there’s no standardised difference from one score to the next, i.e. score in a race or likert scales
Interval Data
Comes in a numerical form and there is standardised differences, for example height
