Anova Data

#1
HI I'm currently using Minitab and want to conduct an Anova test on the pass rate between Classroom based students and Online students. I have carried out anova test before on paint samples hardness, but i'm struggling to carry out this student comparison. I have age, gender and result data of 100 students. how do i carry our the test.

Any advice would be helpful.

Thanks Anthony
 
#3
They are Quantitative, my plan is to do a two way anova, Didn't explain myself very well.

I have two groups, Online learners and campus learners. I have there results and there genders. How do i group these to get an effective result.

The main groups are Online learners and campus learners, that what i want to compare. Should i compare males against males and females against females.

Have two tables of results to discuss?

Thanks
 

Miner

TS Contributor
#4
There are several ways that you can structure this depending on what variables that you want to control for.
  1. You could use a 1-way ANOVA on learning mode. This is essentially the same as a 2 sample t-test. However, age and gender will be part of the experimental noise within groups. This can both obscure the between group differences as well as hide any interaction between learning mode and gender.
  2. You can use a 2-way ANOVA on learning mode and gender. This will isolate any gender differences as well as any interaction between gender and learning mode.
  3. You can use a 2-way ANCOVA on learning mode and gender with age as a covariate. This does everything stated in option 2 as well as isolating age from the within group variation.
Option 3 has the best chance of detecting differences on learning mode as well as identifying possible gender differences and age effect.
 
#5
There are several ways that you can structure this depending on what variables that you want to control for.
  1. You could use a 1-way ANOVA on learning mode. This is essentially the same as a 2 sample t-test. However, age and gender will be part of the experimental noise within groups. This can both obscure the between group differences as well as hide any interaction between learning mode and gender.
  2. You can use a 2-way ANOVA on learning mode and gender. This will isolate any gender differences as well as any interaction between gender and learning mode.
  3. You can use a 2-way ANCOVA on learning mode and gender with age as a covariate. This does everything stated in option 2 as well as isolating age from the within group variation.
Option 3 has the best chance of detecting differences on learning mode as well as identifying possible gender differences and age effect.
Excellent suggestions, number 3 i would like to do on minitab, i'm struggling with how to group them.
 
#7
In Minitab 18, Stat > ANOVA > General Linear Model. Learning mode and Gender go into Factors, and Age goes into Covariates.
The learning modes. I have Online Male, Online Female, Campus based Male and Campus based Female, ages and grades for each group. Is this over complicating things or can i go ahead and with these groupings
 

ondansetron

TS Contributor
#10
It sounds like, currently, you have a variable (let's call it "A" for now) with 4 options: Campus Male, Campus Female, Online Male, Online Female.

I'm suggesting you "decompose" this variable into it's components of Gender and Mode of Learning (call it "Mode").

For the "Gender" variable, tell minitab to look at "A" and for all entries that say "Male" (either campus male or online male) enter "Male" for "Gender" variable (then do the same for female). Then for "Mode" tell minitab to look at "A" and for all entries with "Online" enter "online" for "Mode" (then do the same for Campus). Now you have the same information represented by the separate variables of gender and mode.