ANOVAS

What is the Bonferroni t-test

"Dunns multiple comparison" Protects against type 1 error Uses Bonferroni Correction

The statistical designation for the ANOVA is the ____ statistic

"F" Statistic

What is important about the Bonferroni

-Ahead of time, you believe that there is going to be a difference, before you do an analysis. -You are not as much concerned if there is an OVERALL difference, but instead you believe there will be a difference between groups. Example: If your groups are age groups and the question is .. "Is there a difference in strength between the age groups?" Would you hypothesis ahead of time that there might be some definite in-between group differences? YES. 70 YO should be different from the 20 YO in strength. You can do a bonferroni for this, even if the ANOVA overall is non significant. The anova will look at the whole thing and may not be significant. Where as the bonferroni will analyze the difference between subsets.

Friedmans ANOVA

1 group of sample measured over time. Non-parametric, 1 dependent, more than 2 related samples.

Discuss the interpretation for the One-way ANOVA "F-test" What is tricky about this ?

1. Comes up with 1 statistic which will answer the question "Is there a significant different" 2. If there is a difference, it doesn't tell you where the difference is. Just that there is one.

What are the characteristics of a One-way ANOVA?

1. Only 1 IV 2. One-way, Between groups 3. Continuous data (like an independent T test but more than 2 groups)

The MANOVA is used for _____ dependent variables, And It will result in _______ _____ values

1. Used with multiple dependent variables 2. Will result in several F values 1 F for overall result, then for each dependent variable (using 1 way ANOVA's)

A Two-way ANOVA has ___ IV variables

2 IV's Example, separated by GENDER and then Assigning them to certain GROUPS

In a One-way ANOVA you have to have more than ___ groups

2 groups (you can have as many groups as possible, but it must be more than 2, otherwise you would use an Independent T test) SO it would be continuous data, asking is there a difference between groups (with more than 2 groups) Example: Is there a difference between strength for 3 groups.

A three-way ANOVA has ____ IV

3 Independent variables

2 way and 3 way ANOVAs, main effect vs interaction effect

Main effects: F for each independent variable Ex: "F for gender and F for group" 3rd "F" called interaction effect

But we only want the "interaction F" for 2 and 3 way ANOVAs if its a

Mixed Model (Has repeated measures)

Newmans-Keuls

More powerful, but increases chance for Type 1 error

Scheffe

Most conservative, best with small samples, Protects against type 1 error, But because its conservative it may increase type 2 error

What does MANOVA stand for

Multivariate Analysis of Variance

What do you have to check before running an ANCOVA

Need to check for a linear relationship between variables, if there is a strong relationship, then you can run a co-variate

Will the post- HOC test include the entire sample size

No, ex. 3 groups of 10.. 30 people but only comparing 20 at a time

ANOVAs are identified by the

Number of Independent Variables, 1 IV = One-way ANOVA

A one-way repeated measure ANOVA would test the

ONE GROUP being measured over time, but they are measured mored than 2 times (ex. time 1, time 2, time 3) Anything more than 2 times becomes a "One-way REPEATED measures ANOVA" With TIME as the IV. (similar to paired T-test, but used when more than 2 tests)

This question " Is there a difference between strength in 4 groups" would be analyzed using

One-way ANOVA

A One-way ANOVA is automatically analyzed _____ groups

One-way ANOVA is analyzed automatically between groups, bc we can't add within groups because then it would be a repeated measure.

This question "Is there a difference from admission through D.C for the sample population for strength." would be analyzed using

One-way repeated measures ANOVA. 1 group - "the sample population" 1 DV- "Strength" I IV "time" Repeated measures- " from Admission to D/C)

What are the different types of ANOVAs

One-way, Two-way, Three-way

ANOVAs are primarily a _____ type of statistic

Parametric

If our "One way only F test" is significant then we must complete a

Post-HOC test to determine where the differences are, in essence individual comparisons

Anytime there is the word "response" such as in " Is there a difference in response to treatment between the 3 groups" you know there is a

Pretest / Posttest situation. so a Repeated measures situation

example of a Bonferroni correction with 3 tests

Probability/number of individual comparisons So, Example: 3 tests, .05/3 = α 0.016

A one-way ANOVA can also have ____ measures

Repeated measures, so we can have a One-way repeated measures ANOVA

Whenever you see the word " ____ " and there is more than 2 ____. Then you know its a 2 or 3 way ANOVA

Seeing the word "By" and having more than 2 IV. "is there a Difference between genders "BY" group for strength

How many dependent variables can there be in an ANOVA

Still only 1 DV

ANOVAs are calculated using

Sums of squares called within and between

If a difference is identified in a REPEATED measures ANOVA then you have to do a

T test between samples to see where the difference is, and you will use the Bonferroni correction to alter the criterion alpha to reduce your chance of TYPE 1 ERROR) With the paired samples T-test you will then use the new criterion alpha to determine significance

What is the next step of the MANOVA analysis, that is done by the computer

The computer runs a 1 way ANOVA for each question. ( Is there a difference between question 1 between nursing, OT, PT. If there is a difference then you would do a Post-HOC, is the difference between nursing and PT or PT and OT etc.)

When you do a repeated measure ANOVA, Post-HOC only identifies the ____ not _____

When you do a repeated measure ANOVA Post-Hoc only identifies differences between groups, not within.

When do you use this Bonferroni correction test

With repeated measures ANOVA, You must do multiple paired samples T tests, A Bonferroni Correction is necessary to correct the alpha criteria

A mixed model ANOVA will have both a____ and _____ analysis

Within and Between Analysis. Example: 3 groups: Group 1, Group 2, Group 3 and Each group having a pre-test post-test comparison. Comparing: Response within groups: Group 1 - pretest- vs posttest Group 2 - pretest vs posttest Group 2 - pretest vs postest Between Groups Group 1 versus Group 2 Group 1 versus Group 3 Etc.

In a One-way REPEATED measures ANOVA the analysis is

Within groups versus between groups

The question"Is there a difference in response to treatment between the three groups" would be analyzed by

a Mixed model ANOVA

Tukeys

a Sensitive test that protects against Type 1 error. Good Choice!

A MANOVA is the only analysis of variance where you will have

more than 1 dependent variable

The post HOC tests are a _____ type of statistic

non-parametric

If a difference is identified in a One-way ANOVA you have to use the

post HOC test to determine where the difference is

What is the Bonferroni Correction

probability/number of individual comparisons

In a 3 way ANOVA if any of the IV are time then it becomes a

repeated measures ANOVA

2 way and 3 way ANOVAs will have "Fs" for

Each independent variables, as well as for the interaction

What are the 2 variations of ANOVA that we have

ANCOVA MANOVA

ANOVAs have more than _____, ______ or ______ They give you a ___statistic. If there is significance between groups we need to _____ testing. If there is significance over time (REPEATED measures) then we have to go in and do the multiple ______ _____ and alter the criterion alpha with the _____ _______

ANOVAs have more than 2 groups, 2 tests, or more than 1 IV. They give you a F statistic. If there is significance between groups we need to do Post-HOC testing. If there is significance over time (REPEATED measures) then we have to go in and do the multiple comparison T-test and alter the criterion alpha with the Bonferroni Correction.

ANCOVA

Analysis of Covariance Purpose - determine difference while adjusting for pre-existing variable that may influence results

What does ANOVA stand for ?

Analysis of variance

Summary of how the MANOVA works

Answers the question is there a difference overall between the groups (MANOVA part 1) and then it will say is there a difference for question 1 or question 2 (1-way ANOVA) (done automatically by computer). If you find a significance with the One-way ANOVA then you will do the Post-Hoc thingy.

Kruskal-Wallis (H)

Between groups ANOVA of more than 2 groups. Non-parametric, 1 dependent, > 2 groups

Bonferroni Test versus Bonferroni Correction

Bonferroni test: between groups when you believe there is a difference ahead of time Bonferroni correction: Used for REPEATED measures ANOVA for a paired sample t test.

So a 3 way ANOVA may separate the Data by

Gender, Group, and time > Pre-test Post-test (unlike the 2 way was just Gender and Group or Group and time if repeated measures) This has 3 components

When do we need to use ANOVAs

If we have more than 2 groups/tests or more than ONE Independent variable

When you have a survey that has 10 questions and you want to know if there is a difference you would use the

MANOVA Put in responses for all 10 question.. is there a difference? And then compare to OT group is there a difference for the total responses on this survey between groups?

The Post-HOC test does _____ comparisons

Independent

What would the research question look like for the previous Two-way ANOVA

Is there a DIFFERENCE between GENDERS by GROUP for strength? (since it has more than 2 IV it is not a T-test)

What are the 2 non-parametric ANOVAs

Kruskal-Wallis (H) Friedman's ANOVA

What do they usually choose for probability with the ANOVAs

They typically set it at 0.05 as with the T-test (saying there is a 5% chance that the results are due to chance)

In the mixed model ANOVA ____ is one of the independent variables

Time

What are the 3 different types of Post HOC test

Tukeys, Newmans Keuls, Scheffe,

A variation of the Two-way ANOVA is the ______ which is also called the ______

Two-way repeated measurers ANOVA. Also called the "Mixed model ANOVA" (this is different from the one-way repeated measures because you will have more than 1 group. Having more than 1 group allows you to analyze difference within and between groups. IV become TIME and GROUP)