Finding the average number correct from grid coordinates

Question

I am trying to calculate the average number participants scored correct on a memory task. I have a column called RecallType which tells me if participants were assessed through forward memory recall (called forwards) or through backwards memory recall (called backwards). I also have a column called ProbeState which identifies the type of memory task, of which there are two. In this column I have positions and digits. These are all my variables of interest.

The memory task itself is split by two columns. Recall.CRESP is a column specifying the correct answers on a memory test selected through grid coordinates. Recall.RESP shows participants response.

These columns look something like this:

|Recall.CRESP                     | Recall.RESP                     |
|---------------------------------|---------------------------------|                 
|grid35grid51grid12grid43grid54   | grid35grid51grid12grid43grid54  |                
|grid11gird42gird22grid51grid32   | grid11gird15gird55grid42grid32  |

So for example in row 1 of this table, the participant got 5/5 correct as the grid coordinates of Recall.CRESP matches with Recall.RESP. However in row 2, the participant only got 2/5 correct as only the first and the last grid coordinate are identical. The order of the coordinates must match to be correct.

Ideally I would love to learn from any response. If you do reply please kindly put some comments.

Thanks.

Is your question how to count the number of correct responses? The answer should be 5 for row 1 and 2 for row 2? — Divi, Jan 08 '18 at 18:18
I suppose that would be a better way to word it! Count the correct responses yes, and give the mean by variable of interest. Forwards digit recall is on average eg. 4 out of 5 correct across rows. — Rstudent, Jan 08 '18 at 18:28

score 1 · Answer 1 · answered Jan 08 '18 at 18:46

As you are new to stackoverflow, please read the answer here on how to make a reproducible example so your question is clear: How to make a great R reproducible example?.

From what I understand, you are looking to split your string and then count the equal cases. Some code to get you started on that is below:

a = "grid11gird42gird22grid51grid32"
b = "grid11gird15gird55grid42grid32"

a1 = strsplit(a, "grid|gird")
b1 = strsplit(b, "grid|gird")
table(unlist(a1) == unlist(b1))["TRUE"] - 1

You should be able to take mean by your variable of interest using group_by and summarize functionality of package dplyr.

score 1 · Answer 2 · answered Jan 08 '18 at 20:57

Try using regmatches

fun=function(x)do.call(rbind,regmatches(x,gregexpr(".*?\\d.",x)))
with(dat,rowSums(fun(Recall.CRESP)==fun(Recall.RESP)))
[1] 5 2

DATA:

structure(list(Recall.CRESP = c("grid35grid51grid12grid43grid54", 
"grid11grid42grid22grid51grid32"), Recall.RESP = c("grid35grid51grid12grid43grid54", 
"grid11grid15grid55grid42grid32")), .Names = c("Recall.CRESP", 
"Recall.RESP"), row.names = c(NA, -2L), class = "data.frame")

Finding the average number correct from grid coordinates

2 Answers2