3.3 Flashcards
selection bias
selection of a sample in a non-random way so that it is not representative of the whole population
standard deviation
shows how much variation or spread from the mean exists to question result validity
statistically significant
unlikely due to chance alone
t-test
statistical test that;s useful to determine whether two sets of data are significantly different enough to make a valid conclusion
p-value
produced form a t-test, if value less than 0.05 the null hypothesis can be rejected this the treatment had an effect
to ensure validity and reliability
experiment should address intended aim, test hypothesis, and have confounding factors controlled
confidence and error bars
indicate reliability of data around a mean or percentage chance that the true mean lie somewhere along the bar