T-Test over colums/rows

[tobi-wan kenobi]

Hey guys!

I'm on my report for Riccardos class and I have the following problem: In my experiment, participants went through various conditions at the same time, not through only a certain condition, like it would have been the case in the Lego report or most of our other experiments right. So my data looks kinda like this here:

PhysicalDistanceBuyImPolite EmotionalDistanceBuyImPolite EstimatedTimeBuyImPolite HeatBuyImPolite PhysicalDistanceProfPolite EmotionalDistanceProfPolite EstimatedTimeProfPolite HeatProfPolite


(v1) (v2) (v3) (v4) (v5) (v6) (v7) (v8)

19 89 35 67 21 69 16 56
39 100 67 56 42 69 66 51

50 100 100 66
71 97 76 28 77 71 80 68
25 30 75 73 34 38 60 59




19 100 52 70 50 78 51 45
10 100 61 60 22 75 50 49
67 100 83 50 60 70 50 50
73 100 17 60 68 80 28 32

The point is, that each of the block of columns (v1 to v4; and v5 to v8) already contains the information on the condition, because the conditions are just the situations they went through. That is: In order to compare the "polite" with the "impolite condition" i cannot compare a row with a colum (like i would normally do) but i have to compare two colums one as the function of the other, because they already contain the info on which condition they were in. and i cant add that info to my dataset because then I would need a third dimension to measure the variables against.

In fact the whole thing is even more fucked up because: I actually have three conditions and would like not only to regard them as different factors but ALSO as numerical dimensions (increasing politeness from "rude" over "polite" to "overly polite"). I have tried to convert my dataset in various forms but i always and up with the problem that the Rows as such are already indicators of the condition and so i cannot make one variable the function of another variable (since there is no such thing as a "condition" variable).

HELP PLEASE!

Thanks a lot in advance... I hope my formulation makes sense and the problem is clear

[tobi-wan kenobi]

So guys!

I found a solution, so if you ever have the same problem just do what I did. I just rearranged the data in the following manner. I collapsed (not in R but manually) the colums v1 + v5; v2 + v6; v3 + v7 & v4 + v8. that is i just allocated the latter beneath the former. To conserve the informative value of the thing I created a new colum that contains the info on the condition so that I now have this:

V1 v2 v3 v4 cond

19 89 35 67 1
39 100 67 56 1

50 100 100 66 1
71 97 76 28 1
25 30 75 73 1




19 100 52 70 1
10 100 61 60 1
67 100 83 50 1
73 100 17 60 1
21 69 16 56 2
42 69 66 51 2

50 100 100 66 2
77 71 80 68 2
34 38 60 59 2




50 78 51 45 2
22 75 50 49 2
60 70 50 50 2
68 80 28 32 2

and now i can use shit the way i want in R =) Sorry for the confusion i hope this method might still help you