Three researchers ran subjects for this experiment. Each ran a different number of subjects and different kinds of models, as tallied below:
##
## concave convex
## Bernice 1 1
## Dietmar 2 3
## Laura 4 3
Researchers tended to take different amounts of time when running the conditions. This may be influenced by low n.Â
What is happening during the time differences between researchers? Are the different researchers also speaking different amounts during each of the conditions?
Note: should revisit the time speaking transcripts to make sure there aren’t “hallucinations” in it. The earlier transcripts sometimes filled the model cleaning time with halluncinated nonsense.
Are the subjects speaking different amounts for each of the researchers?
(Note: same as above, check these for halluncinations)
Do the different condition times by different researchers relate to different touch behaviors? Did particular researchers seem to elicit more touching of the model, either by touch ratio or time?
Something weird is going on for condition 1 Bernice–we seem to be missing a data point. But, overall it doesn’t seem like different researchers elicited different touch ratios. There may be small difference in touch time by researcher.
What if we look across conditions?
The subjects spent varying amounts of time with the model for each condition in the experiment:
## $`1`
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 3.841 6.251 8.157 7.810 9.135 12.686
##
## $`2`
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 4.204 6.941 8.898 9.193 10.829 19.347
##
## $`3`
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 11.15 15.41 18.80 18.45 21.96 25.75
The amount of time spent within each condition did not seem to depend (strongly?) on whether the model was convex or concave.
Different subjects spent different amounts of time with the model for each condition.
One might guess that spending more time with the model would lead to more words spoken or more time speaking (either total, for the subject, or for the researcher). Word counts and time speaking for the researcher:
Word counts and time speaking for the subject:
Seems like a good fit for both cases–so, generally, more time spent with the model corresponded to more time talking and more words spoken for both the subject and the interviewer. [[to do: could add fit lines to this]]
One might guess that spending more time with the model would lead to more area touched by the subject or more time touching by the subject. Area touched:
…area touched by the subject does not seem to correlate with condition time.
Time touching by the subject:
Maybe a little bit of correlation between condition time and time spent touching (hard to say, mushy graph). Doesn’t seem to be correlation between condition time and percent of time spent touching during the condition.
Touch and talking–amounts and patterns of time of: (1) subject touching + talking (dark blue); (2) subject touching + researcher talking (dark red); (3) no touching + subject talking (light blue); (4) no touching + researcher talking (light red); (5) subject touching + no talking (grey); (6) nothing happening (offwhite)
The same plots by subject and condition
How is time is spent?
How is time is spent by condition?
Touch behavior on an individual basis–are there people who are high touch and people who are low touch? or something else more detailed? consider time as well
For more details on using R Markdown see http://rmarkdown.rstudio.com. When you click the Knit button a document will be generated that includes both content as well as the output of any embedded R code chunks within the document.