7118 Lecture Notes - Lecture 3: Repeated Measures Design, Null Hypothesis, Analysis Of Variance
20 Jun 2018
School
Department
Course
Professor
LECTURE 3A
One-way ANOVA
ANOVA: Basic concepts
- Multiple groups
oWant to compare more than 2 groups
oEg timepoints
Before semester
During semester
After smester
oEg environments
High stress
Low stress
Medium stress
- ANOVA
oAnalysis Of Variance
oUsed when comparing more than 2 groups
oOne way between groups ANOVA: one IV with multiple levels/groups that was
manipulated between subjects
oOne way repeated measures ANOVA: one IV with multiple levels/groups that was
manipulated within subjects
oBoth involve IV with several groups, and one dependent variable factorial ANOVA is
when you have more than one IV
- Things we already know
oStart with null and alternative hypothesis
oCalculate the test statistic (F – not t or z)
oFind out what the probability of getting this result if null hypothesis is true
oIf p is smaller than alpha then reject null hypothesis
oCan calculate the effect size, type 1 error, type 2 error and power
- Whats new
oCalculate F instead of t
o2 degrees of freedom (df)
oFormulae are different
oNeed extra step at the end to find out where the difference is
First find if there is a difference across all of the groups
Then (if there is a significant difference) run extra tests to find out where the
difference is
- The structural Model
find more resources at oneclass.com
find more resources at oneclass.com
o
o
o
oFor me
19 = 26 + 11 + (-18)
oHeight for me
165 = 160 -10 + 15
find more resources at oneclass.com
find more resources at oneclass.com
o
- One-way between groups ANOVA
oWhen one IV, with more than 2 groups, and manipulated between subjects
oOne way means ‘one independent variable’
o
IV is condition, with 5 levels
DV is number of words remembered
oSide notes
Could run a lot of t tests
But because we are running lots of tests with the same set of data, there is a
higher risk of making a type 1 error
The ANOVA checks for us whether there are any differences anywhere between
the groups – if so we then conduct follow-up tests to find out exactly what those
differences are (between which groups)
oHypotheses
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Multiple groups: want to compare more than 2 groups, eg timepoints. Whats new: calculate f instead of t, 2 degrees of freedom (df, formulae are different, need extra step at the end to find out where the difference is. First find if there is a difference across all of the groups. Then (if there is a significant difference) run extra tests to find out where the difference is. The structural model o o o: for me. 19 = 26 + 11 + (-18: height for me. One-way between groups anova: when one iv, with more than 2 groups, and manipulated between subjects, one way means one independent variable" o. Dv is number of words remembered: side notes. But because we are running lots of tests with the same set of data, there is a higher risk of making a type 1 error.
