spss Flashcards
(21 cards)
When do you use correlation tests?
If you are interested in the degree of relationship or association between two (or more) variables
If the date is normally distributed what correlation test is used?
Pearson’s
If the date is not normally distributed what correlation test is used?
Spearman’s
How do you get the Coefficient of determination(R2) in correlation testing?
square the correlation value
What does The correlation value show?
shows the strength (i.e., small, medium, large) and direction (i.e., positive or negative) of the relationship
What does The ‘Sig’ value show?
shows if the correlation is ‘significant’ or not also known as P value.
How do we calculate skewness and kurtosis values?
dividing them by their standard error (i.e. calculate ZSkewness and Zkurtosis), then assess them against the +/- 1.96 range.
How do I perform skewness and kurtosis tests?
Analyse > Descriptives > Explore
When do I use regression analysis?
If you are interested in establishing whether a predictor variable (simple regression), or predictor variables (multiple regression), predict a criterion variable.
How do I conduct regression analysis when data is normally distributed?
Analyse > Regression > Linear
How do I conduct regression analysis when data is not normally distributed?
bootstrap
dependent?
(outcome/effect)
independent?
(manipulated/cause)
In a simple linear regression, how do you find the correlation value?and what does that mean?
referred to as ‘r’ and shown as ‘R’ in SPSS, which
demonstrates the strength (i.e., small, medium, large) and direction (i.e., positive or negative)
In a simple linear regression, how do you find the Coefficient of determination?and what does that mean?
R2 and shown as ‘R Square’ in SPSS,
which demonstrates the variation in the dependent variable explained by the independent variable
What does The ‘ANOVA’ box show\?
tells you whether the model, overall, results in a significantly good degree of prediction of the outcome
variable
What does The ‘Coefficients’ box show?
tells us if each independent variable is a significant predictor of the dependent variable (again, we look at the ‘Sig’ value to establish this).
What does this equation stand for, Y = a + (b*x)
Y = outcome measure a = intercept coefficient b = slope coefficient x = individual predictor variable
When do I use difference tests?
If you are interested in establishing if there is difference
How do I conduct an independent t-test using SPSS?
Analyse > Compare Means > Independent Samples T-Test
How do you calculate effect size?
Cohen’s d. You can do this by entering the mean, standard deviation and number of participants (n)
