Between listener convergence pilot data

There are 7 different .csv files. Due to the way multi-participant experiments are run on Labvanced, each experimental Condition is recorded in its own dataset.
The three ending in -utf8 are the raw datasets, the ones that end in -clean are the cleaned ones with the relevant information. Turning the datasets into a single one is easy enough, if you wish to do so, I personally like to keep them separate until I absolutely have to combine them.

The three conditions are as follows:

fb: feedback – participants were incentivized to agree
falsefb: false feedback –participants were incentivized to disagree
nofb: no feedback – individual control condition without any interaction.
The cleaned data in the two interactive conditions has the following variables:

Pair_ID: identifies participants in a dyad
Condition_Id: Stimulus level (1 corresponds to /s/, 5 to /S/)
Stimulus_order: first iteration of a given stimulus to the last (i.e. „13“ means the 13th time a participant has heard the corresponding acoustic stimulus in the Condition_Id column).
decision1: picture chosen by participant 1 in a dyad
decision2 picture chosen by participant 2 in a dyad - 3 here corresponds to picture associated with /s/ in the training, 2 corresponds to the picture associated with /S/. (I don’t know at which point in the cleaning script the strings were replaced with numbers, but numbers work as well, I think).
agreement: 1=participants agree on categorization of the stimulus,
0 = participants disagree on the categorization.
Trial_Nr: Order in which the acoustic stimuli were presented (randomized)

Th individual condition has of course only one decision column and no agreement, individual participants are identified as subj_counter_global.

the .csv file called „identall“ shows how the overall identification of the stimuli differs by condition.
The relevant Variables are the following: Condition_Id: Acoustic stimulus from 1(/s/) to 5 (/S/)
n: Number of times participants within a condition chose picture 1 for a given stimulus identperc: n transformed into percent task: Experimental condition.

the other columns in identall are a result of my data cleaning process and can be disregarded
I am also uploading my annotated cleaning and plotting script.