descriptive and inferential statistics

what are descriptive statistic used for in research

summarize data

how do they help summarize data before inferential testing

allows researchers to make predictions or inferences about a population based on a sample

negative skewed

median > mean

positive skewed

median < mean

which measure of central tendency is more accurate when data is skewed

median

which measure of central tendency is more accurate when data is noramlly distributed

mean

floor effects

poorly designed measure that causes all scores to pool toward the lowest values

ceiling effect

poorly deisgned measire that causes all scores to pool toward the highest value

what is the purpose of inferential statistics

allows us to infer that the results from our sample can apply to a larger population

how do inferential tests tell us whether differences or relationships are statistically significant

demographics, baseline variables, outcome variables

what do p values tell us about the role of chance

probability that an observed difference could have occured just by random chance

what does alpha represent

probability of rejecting the null hypothesis when its true

when do we say a result is statistically significant

a lower p value, if a p value is less than the alpha value

what determines which statistical test should be used

test of assocaition and test of differences

what are parametric test and what assumptions must be met

allows for greatest statistical power but assumptions have to be met: continuous variables, adequate sample size, normal distribution, variance is the same in both group, all measures within agroup are independent

what are non parrametric test and when should they be used

more prone to type II errors bu greater flexibility in terms of assumptions