PSYC30013 Lecture Notes - Lecture 9: Time Point, Repeated Measures Design, Analysis Of Variance
18 Jun 2018
School
Department
Course
Professor
Lecture 9
Now consider the extension of the dependent two-group design to research
questions involving three or more groups...often called repeated measures ANOVA
designs.
Major issues for one-way between-subjects ANOVA also apply:
1) The use, meaning, and value of the omnibus approach.
2) Decomposition of total sum of squares into meaningful additive component sum of
squares.
3) Use of planned comparisons as the most appropriate and preferred method for
directly investigating within-subject group differences.
The ways in which these issues are addressed, however, differ in general for within-
subjects designs.
With dependent two-group designs, we noted that dependency came about by
measuring. . .
→ Same group of people at two different time points, or
→ Paired individuals in matched groups.
on the same DV.
With three of more groups, a matched design is very rare (unless, e.g., research was
being undertaken into triplets).
Within-subjects designs virtually always reflect repeated measurements of the same
group of people at three or more time points—we only consider this kind of design.
An important assumption at the outset for within-subjects ANOVA using repeated
measures is that the interval between two adjacent time points is the same for all
persons.
That is, if we have, e.g., 3 time points in the design then. . .
⇒ Duration between the first and second time point should be the same for all
persons, and
⇒ Duration between the second and third time should be the same for all persons as
well.
This does not mean that durations between adjacent time points must be the same.
→ e.g., the first duration may be one month, whereas the second duration between
the 2nd and 3rd time point may be 3 months.
→ Note that, in practice, slight variations in duration will be of no major consequence)
find more resources at oneclass.com
find more resources at oneclass.com
find more resources at oneclass.com
find more resources at oneclass.com
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Now consider the extension of the dependent two-group design to research questions involving three or more groupsoften called repeated measures anova designs. The ways in which these issues are addressed, however, differ in general for within- subjects designs. With dependent two-group designs, we noted that dependency came about by measuring. Same group of people at two different time points, or. With three of more groups, a matched design is very rare (unless, e. g. , research was on the same dv. being undertaken into triplets). Within-subjects designs virtually always reflect repeated measurements of the same group of people at three or more time points we only consider this kind of design. An important assumption at the outset for within-subjects anova using repeated measures is that the interval between two adjacent time points is the same for all persons. That is, if we have, e. g. , 3 time points in the design then.
