273 E4: ANOVAs

>2 levels of the IV

according to the decision tree, what is the first branch that determines if the test is an ANOVA?

omnibus

an _______ test asks: is there a difference across ALL conditions or across ALL time points that is greater than would be expected due to chance?

There is not a difference across all groups

what is the null hypothesis of an ANOVA?

t-test

based on the null hypothesis alone, what does an ANOVA null also mimic?

increases

what happens to type I error rate when we preform multiple tests?

.01, 0.06, 0.14

what are the small, medium and large values of peta square?

bonferroni post hoc

if you REJECT the null, what test do you also need to run?

Levene’s

running the _______ test tells you if the variation is equal among groups

We fail to reject the null hypothesis, [F(dof) = num, p = 0.227, peta =), there is no differences. The assumption of equality of variances was met (levenes p).

suppose you have a p = 0.227. how would you report the findings?

mixed ANOVA

scenario: you have 2 groups: control + experimental, and 2 times (pre and post). what test do you apply?

interaction

a graph of a mixed anova is interested in the _______ of the two group responses

one way anova

What test is performed? Scenario: A sports medicine researcher wants to know whether the number of previous concussions differs depending on which sport an athlete plays (soccer, football, or lacrosse).

We reject the null hypothesis, F(2, 27) = 4.460, p =0.021, peta = 0.24… run post hoc

What is the first part of the statement, and how should you preceed?

A sports medicine researcher wants to know whether the number of previous concussions differs depending on which sport an athlete plays (soccer, football, or lacrosse).

Output: df: 2, 27, F = 4.460, p = 0.021, peta = 0.24.

Football players saw more concussions than soccer players (p=0.018), there was no difference in concussion in football and lax (p = 0.525), or lax and soccer (p = 0.369).

Write the second half of the report statement:

A sports medicine researcher wants to know whether the number of previous concussions differs depending on which sport an athlete plays (soccer, football, or lacrosse).

post Hoc:

football - lax: 0.525

football - soccer: 0.018

lax - soccer: 0.369

1 way repeated measures

What test is applied for the given scenario?

The team physician wants to test whether symptom severity changes from before, to after a mild traumatic brain injury (mTBI), and through recovery in this sample.

We reject the null hypothesis, (F(2, 58) = 34.755, p <0.001, peta = 0.545], bonferoni

Write the first part of the null based on the output, what should be run based on this conclusion?

The team physician wants to test whether symptom severity changes from before, to after a mild traumatic brain injury (mTBI), and through recovery in this sample.

df = 2, 58, F = 34.755, p <0.001, peta = 0.545

Symptoms increased from pre to post (p<0.001), symptoms increased from pre to recovered (p =0.002), and symptoms decreased from post to recovered (<0.001).

The team physician wants to test whether symptom severity changes from before, to after a mild traumatic brain injury (mTBI), and through recovery in this sample.

Post Hoc:

Pre-post: <0.001, mean difference = -3,8

Pre-recovered: 0.002, mean differecne - -0.767

post-recovered: < 0.001: mean difference = +3.1

mixed ANOVA

Scenario: what test is applied?

An athletic trainer wants to know whether cognitive performance is affected by a mild TBI when athletes are compared to their teammates (others playing the same sport but without a mild TBI).

We reject the null hypothesis, F(num) = 250, p <0.001, peta = 0.899, Bonferoni

write the first part of the statement, and how should you proceed?

An athletic trainer wants to know whether cognitive performance is affected by a mild TBI when athletes are compared to their teammates (others playing the same sport but without a mild TBI).

Output:

Time * Group: F = 250, p <0.001, peta = 0.899

1 way repeated measures anova

Scenario: what test should be applied?

Now the athletic trainer wants to know whether changes in cognitive performance after a mild TBI from pre- to post-injury differs between athletes playing different sports.

We fail to reject the null hypothesis [F(num) = 0.497, p = 0.620, peta = 0.076, there is no sport time interaction. Time is a main effect [F(num) = 376, p <0.001, peta = 0.969]. Sport is not a main effect [F(num) = 0.926, p = 0.0423, peta = 0.134.

What is the conclusion?

Now the athletic trainer wants to know whether changes in cognitive performance after a mild TBI from pre- to post-injury differs between athletes playing different sports.

[Within subjects]

time: F = 376, p <0.001, peta = 0.969

time * sport: F = 0.497, p = 0.620, peta = 0.076

[between subjects]

Sport: F - 0.926, p = 0.423, peta = 0.134