stats

null hypothesis definition

hypothesis that assumes no difference, no effect, or no correlation.

what does standard deviation show

how much variation there is between data and mean value

what does high/low standard deviation mean

high → lots of variation, not very reliable

low → little variation, more reliable

standard deviation equation

what does spearman’s rank correlation assess

the degree of association between measurements from the same sample

spearman’s rank correlation equation

• the closer rs is to 1+ or 1-, the more likely the correlation

does spearman’s rank correlation need a null hypothesis

• yes

• degrees of freedom for critical value = n

• if rs is closer to 1+ or 1- than the critical value, reject the null hypothesis

what does chi square determine

if there is a connection between discrete data

chi square equation

does chi square need a null hypothesis

• yes (there is no significant difference between observed and expected)

• degrees of freedom for critical value = n - 1

• t ≥ CV reject the null hypothesis

• t < CV accept the null hypothesis

what does the paired t-test assess

if there is significant difference between data sets with two categorical groups with the same individuals

paired t-test equation

does the paired t-test need a null hypothesis

• yes

• degrees of freedom for critical value = n - 1

• t ≥ CV reject the null hypothesis

• t < CV accept the null hypothesis

what does the unpaired t-test assess

if there is significant difference between data sets with two categorical groups with different individuals

unpaired t-test equation

does the unpaired t-test need a null hypothesis

• yes

• degrees of freedom for critical value = (n1 + n2) - 2

• t ≥ CV reject the null hypothesis

• t < CV accept the null hypothesis

what test do you use - looking at frequencies, and whether observations differ from expected values

chi sqaured

what test do you use - looking at the relationship between two variables

spearman’s rank correlation

what test do you use - looking at whether there is a difference in the mean between the same group before and after a change

paired t-test

what test do you use - looking at whether there is a difference in the means between two independent groups

unpaired t-test