Please enable JavaScript.
Coggle requires JavaScript to display documents.
inferential statistics, survey research - Coggle Diagram
inferential statistics
The Null Hypothesis
-
null hypothesis
- no difference or relationship between parameters in the populations
- any difference or relationship found for the samples is the result of sampling error.
types of test
two- tailed test
-
null hypothesis
- no difference between the groups (A=B)
- two-tailed test allows for the possibility that a difference may occur in either direction (A>B or B>A)
one-tailed test
-
null hypothesis:
- one group is not better than another, if a difference occurs it will be in favour of that particular group (A>B)
Standard Error
sampling error - if a difference is found between two sample means, the important question is whether the difference is a true or significant one or just the result of sampling error.
normal distribution
- most sample means will be close to the population mean
- very few means are much higher or lower than the population mean
- mean of all the sample means – good estimate of the population mean
Test Of Significance
helps on deciding whether to reject the null hypothesis
infer that the difference is significantly greater than that of chance.
-
significance of:
T-Test
-determine whether two means are significantly different at a selected probability/significance level
-look at the p-value, if less than or equal to selected, then there is a statistical difference between the two means
Pretest-Posttest
- use t-test for pretest, if the difference is not significant, use t-test on the posttest
- use t-test for pretest, if the difference is significant, use ANCOVA (analysis of covariance) to adjust posttest scores for initial differences on some variable (in this case the pretest)
-
Chi Square
-
compares the proportions actually observed in a study to the proportions expected, to see if they are significantly different
-
-
simple (One-way) ANOVA
Analysis of Variance
determine whether there is a significant difference between two or more means at a selected probability level
look at the probability level associated with F statistic, if less than the preselected, then there the difference is statistically significant
-
-
survey research
-
-
-
descriptive research
- typical survey studies that assess attitudes, opinions, preferences, demographics, practices, and procedures
-
-
-