Stats Flashcards
Chi - squared test
Test significance of difference between observed and expected results
Students t - test
Test the significance of difference between means
Correlation Coefficient
Test significance of correlation (positive or negative) between two variables
Null Hypothesis
No significant difference
When do we reject null hypothesis?
When p is smaller than or equal to 0.05
There is a significance difference/correlation
For chi squared rejection of null hypothesis requires
A value for chi - squared that is greater than the critical value
Stats conclusion paragraph for chi’s squared
My value is less than critical value hence no statistically significant difference between observed and expected and we accept null hypothesis. Any difference is due to chance alone taken at 5% significance level hence there is a 5% or less probability of difference in results being due to chance.
My value is more than critical value at five percent significance level, hence there is a five percent or less probability of difference in results being due to chance alone. There is a statistically significant difference between observed and expected results, hence reject null hypothesis
When can students t - test be used?
Two categorical groups
Independent data
Interval data (each unit is the same size)
In students t test when can we reject null hypothesis
If t value is greater than critical value
T test paragraphs
Significant difference between means because my vakue is larger than t value hence reject null hypothesis at 5 significance level hence 5% or less probability of difference in means being due to chance.
My value is less than critical value, accept null hypothesis, no significant difference between means at five percent significance level, any difference has 95% or more probability of being due to chance
