Module 17: Hypothesis testing Flashcards
Hypothesis testing
is the process of determining whether a statement is supported by the results of the research project
Null-hypothesis (statistical hypothesis)/H0
this always predicts that there is no difference between the groups being compared. The researcher does not expect to find this, they want to prove this wrong.
Statistical notation: H0 = µ0 = µ1
Alternative hypothesis/H1/Ha
this is where the researcher predicts that there is a significant difference between the groups.
Statistical notation: H1 = µ0 > µ1
One-tailed hypothesis (or directional hypothesis)
an alternative hypothesis in which the researcher predicts the specific direction of the expected difference between the groups
Two-tailed hypothesis (or nondirectional hypothesis)
an alternative hypothesis in which the researcher expects to find differences between the groups but is unsure what the differences are
Type I error
If we reject the null-hypothesis (that there is no IQ difference between groups), we may be correct or incorrect. If our decision to reject H0 is correct, then there truly is a difference in IQ between children in academic afterschool programs and those in the general population. However, our decision could be incorrect. The result may have been due to chance.
Type II error
When we do not reject the null-hypothesis, saying there is no difference between groups when in reality there is. We have missed the difference that really exists and have failed to reject the null-hypothesis when it is false.
Chance
The sample is random, you might by chance alone get a sample that shows a mean difference between these groups
Parametric tests
are tests that require us to make certain assumptions about estimates of population characteristics, or parameters. These assumptions typically involve knowing the mean (µ) and standard deviation (σ) of the population and knowing that the population distribution is normal. Parametric statistics are generally used with interval or ratio data.
Nonparametric tests
are tests that do not involve the use of any population parameters. In other words, µ and σ are not needed, and the underlying distribution does not have to be normal. Nonparametric tests are most often used with ordinal or nominal data.