Stats Flashcards
What is a categorical, ordinal and continuous data?
Categorical-each point assigned to category
Ordinal-in logical order but no info about difference eg race scores
Continuous-ranking and scale. Decimal points eg temp
Measure of central tendency
Best guess about typical score.
Mode, median and mean.
Measures of dispersion
Best guess of spread
Range, variance And standard deviation.
What is distribution
How the typical value and spread of scores relate to each other
Characteristics of normal distribution
Bell shaped, scores spread evenly around mean,
68-95-99.7% rule
Most of population will lie within 3 SD from mean
What is the confidence interval for
With 95% certainty the population mean will lie within
What’s a between subjects design
Diff Participants take part in diff conditions
No practice or Order effects. Less chance of demand characteristics. But need more participants and have individual differences
What’s a within subjects design
Same participants take part in all conditions.
Few participants needed and less individual differences
But practice effects, demand characteristics
What’s the P value for
Needs to be less than 0.05 to be significant. It is the probability that we get th same results due to sampling error if null was true
When do you use an independent t-test
With a between participants design. Comparing means of 2 independent groups.
When do you use a paired t-test
With a within design. Comparing differences between two means of same group
What’s a percentage overlap
If little difference between groups then will overlap
What is levens test for
If not significant then use equal variance assumed.
What is degrees of freedom
The number if scores that are free to vary after final calculation
Independent-n-1+n-1
Correlation n-2
What are type 1 and type 11 error
Type 1-accept null when it’s true
Type 11-reject null when should accept
What is a correlation
Relationship between two continuous variables
What’s correlation coefficient
Strength of relationship Inbetween -1 and 1
What is shared variance and equal variance
Shared variance is r squared. And this is the amount of variance that is explained for. Unique variance is any that isn’t explained. The bigger shared the strong the relationship
What is chi squared used for
Relationship between two categorical variable.
Makes inferences about likelihood
What are the expected and observed values
Expected is what you would presume would be the scores if it was spread out equally. Calculate difference between cells
What are the degrees of freedom for chi squares
(Number of columns-1)x(number of rows-1)
What assumptions should be met for chi-squared
No more than 25% of cells should have expected value less than 5. And no individual cell should have expected value less than 1
What do non parametric tests use
Use ranks and less powerful
What’s the Shapiro wilk test
If not significant then data is normally distributed
What non parametric test is used for independent t-test
Mann Whitney-compares mean ranks of 2 independent groups.
All scores are ranked and calculates how many times one condition is rated higher than other.
Non-parametric for paired t-test
Wilcoxon. Comparing mean ranks of participants who scored higher in on condition than other. Calculate the difference and order and rank. Ignore ties. Separate positives and negatives