# Analysis of variance, help!

hello,

I doing my masterÂ´s thesis and I need some help with statistics. My professor said I should do a analysis of variance of my data. I have tried to understand the books and web-material about the subject but I just donÂ´t get it..

I have a data of 230 survey-answers coded in Excel. The data is about environmental attitudes of boat owners. Now I should find out if there is a difference in attitudes between two groups, sailingboat owners and motorboat owners.

I downloaded Statistixl and it seems that I can do the analysis by using it. One the problems is that I donÂ´t have a clue what kind of results (numbers) I should get. What is a "Y-variables range" "Factor range", "Covariates range"..?

I really hope someone could help me.

best regards
miika j. heikkonen
miheikko@mappi.dontspam.helsinki.fi

• Hi Miika

It's hard for me to answer without knowing more about your data. I would suggest that you have a read through the statistiXL help file, looking especially at the examples for ANOVA to see if your data matches the sort used there. To me it doesn't initially sound like your data is the sort to be analysed by ANOVA and you might want to check out the Discriminant or Factor Analysis sections in the help file.

Cheers

Alan
• I'm having similar difficulty understanding how to carry out an analysis of variance on my data. I have 7 sets of data, and want to analyse them in order to find out if the means of sets 2, 3, 4, 5, 6, and 7 are significantly lower than the mean of set 1 (a benchmark). The sets look like this (but with about 300 rows instead of 4):

Set1 Set2 Set3 Set4 Set5 Set6 Set7
4 7 8 7 3 6 5
7 3 5 5 5 6 5
4 3 5 3 3 4 3
1 2 3 3 1 1 2

This seems to be the same layout as the example data on the "ANOVA Introduction to input" help page which has columns labelled as "Grayson's Pond", "Beaver Lake" etc.
So, when I go to the StatistiXL menu and select "Analysis of Variance" and "Full Factorial", I'm confused about what I should select as the "Y Variable(s) Range", and "Factor(s) Range". If you could give an example of what should be selected for these ranges either in relation to the data above, or the example data shown on the help page, that'd be a great help.

Peter Briggs
• Hi Peter

You can enter your data in 2 ways. One way is in the format you list, with Set 1...Set 7 in the first row and the data underneath. If we start at cell A1, the sample you have provided would extend from A1 to G5.

In this case, if your data is on Sheet1 of your workbook, you would enter Sheet1!$A$1:$G$5 as the Y Variable(s) Range and leave the Factor Range blank.

The second method is to enter each datapoint as a row. In this method you would have the following 2 columns of data

Factor Variable
Set1 4
Set1 7
Set1 4
Set1 1
Set2 7
Set2 3

and so on

Your Factor data would extend from cell A1 (if you include the heading "Factor") to cell A29. Your associated values would extend from cells B1 to B29. With this scenario you would enter Sheet1!$B$1:$B$29 as the Y Variable(s) Range and Sheet1!$A$1:$A$29 as the Factor Range.

I hope this helps

Alan
• Great, that works fine - I was getting confused by the terminology.

Cheers for the help!
Peter
• Okay, I'll try my best not to make this a question for a statistics tutorial, but I might not succeed....
Anyway, I have a 2-factor (I think) nested design, and I can't find the option in the software to perform this, even though it's described in the help file. So I guess I'm wondering if the trial version doesn't have this available, or if I have to do the ANOVA at each level?

Thanks for any help.
• Hi James

The trial version is fully functional. Are you saying you have 2 nexted factors of 2 standard factors plus a nested factor?

Cheers

Alan
• Hello Alan,

The design of my experiment I have detailed here. I'm really not good with the terminology, however I think what it is is that I have two standard factors (infested vs. uninfested), and then nested within them are 3 factors (site, trap and bottle).

And what I could really use some help with is how I should lay out my data in the spreadsheet itself, although Excel makes it rather easy to copy. Anyway, that's beside the point ...