Sample data from a visual-world eyetracking study included to demonstrate use of gmpm functions on multilevel data with difficult dependencies. Data are from a subset of Experiment 2 in Kronmuller and Barr (2007). A parametric reanalysis of these data can also be found in Barr (2008).
A data frame containing 2464 rows, each one an aggregate over multiple trials within the same condition for the same subject.
|[,2]||Speaker||factor||identity of the speaker (same,diff)|
|[,3]||Load||factor||was listener under cognitive load?|
|[,4]||AggID||numeric||index of unique speaker*load*subject combinations|
|[,5]||ms||factor||time in ms from description onset|
|[,6]||t1||numeric||time in 100 ms increments|
|[,7]||T||numeric||count of looks to target|
|[,8]||O||numeric||count of looks to non-target object|
|[,9]||X||numeric||count of frames including looks to blank regions or blinks|
|[,10]||N||numeric||total number for frames (T+O+X)|
|[,11]||NT||numeric||total non-target frames (N-T)|
Dale J. Barr <[email protected]>
Kronmuller, E. and Barr, D. J. (2007). Perspective-free pragmatics: Broken precedents and the recovery-from-preemption hypothesis. Journal of Memory and Language, 56, 436–455.
Barr, D. J. (2008). Analyzing 'visual world' eyetracking data using multilevel logistic regression. Journal of Memory and Language, 59, 457–474.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
data(kb07) # first let's fit the data using a binomial model # kb07.binom.gmpm <- gmpmCreate(cbind(T,NT) ~ t1*Speaker*Load | SubjID, "binomial", kb07, ivars=c("Speaker","Load")) # you will need to increase the number of runs in the command below # (to, say, 999) to get sensible results kb07.binom.gmpm <- gmpmEstimate(kb07.binom.gmpm, list(maxruns=19)) summary(kb07.binom.gmpm) # now let's do a more powerful multinomial analysis # where we break out looks to blank regions and blinks # into a separate "junk" category (X) kb07.mnom.gmpm <- gmpmCreate(cbind(O,T,X) ~ t1*Speaker*Load | SubjID, "multinomial", kb07, ivars=c("Speaker","Load")) # again, you'll need to increase the number of runs to get something # sensible kb07.mnom.gmpm <- gmpmEstimate(kb07.mnom.gmpm, list(maxruns=19)) summary(kb07.mnom.gmpm)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.