Please enable JavaScript.
Coggle requires JavaScript to display documents.
Inferential Statistics (Inferential statistical tests (Parametric and non…
Inferential Statistics
-
Level of significance
Can be used, most common level is P<0.05. So less than 5% probability that results occurred by chance. So weren't due to the IV.
Null Hypothesis
Express results in terms of null hypothesis, result is statistically significant reject the null and accept the alternative, if it isn't statistically significant then we accept the null.
Type 1 error
False positive- Say results are significant when they aren't When find a difference due to chance. Often caused by a too lenient significance level.
Type 2 error
False negative- When we fail to detect a real difference. Say results not significant when they are. Significance level is too strict.
-
-
-
-
-
-
Significance
If calculated score is lower than critical score results are significant, so null is rejected, alternative accept. (Mann-Whitney, Wilcoxon signed ranks and Binomial)
If calculated score is bigger than your critical value you accept the alternative and reject the null (chi squared and spearmans rho).