Specification of the experimental conditions for each trial
The experimental conditions for each trial are defined in the conditions.csv file, the format of which is explained here
The conditions.csv file is used by a few pieces of the JCT software to connect the experimental conditions (such as role assignment and ability to speak) to the specific trials within an experiment. Since the conditions changed between the first and second versions of the experiment, there are some differences between the related csv files.
To generate your own conditions, at the moment, you would need to edit the gdf2nxt.py (in software/gdf2nxt) and the reformatcc.py (in software/JASTAnalyser) scripts and define a unique identifier in your data to specify the conditions file format (we prefix all our filenames and ids with "jcte1" or "jcte2" which we can use for this purpose).
Values in square brackets [] indicate the range of allowed values.
Where a particular experiment/pair/dyad has to be repeated, the related files should be numbered 32 higher than the discarded data if that data is to be kept. The pair recorded in the csv file will be this experiment number modulo 32.
Fields (in left-right order) for experiment 1
- PAIR
- dyad id (as in jcte1_id). [01-32]
- ROLE
- is role assigned? [0 = false, 1 = true]
- SPEECHORD
- is speech in first or second half of dyad? [1|2]
- TANGRAM
- Tangram used for this trial. [1-16]
- ORDER
- Position of this trial within dyad. [1-]
- SPEECH
- is speech allowed in this trial? [0 = false, 1 = true]
- GAZE
- is other eye gaze displayed in this trial? [0 = false, 1 = true]
Fields (in left-right order) for experiment 2
- PAIR
- dyad id (as in jcte1_id). [01-32]
- PART
- N = Part of dyad (1st, 3nd, 3rd), as in jcte2_id_pN, for splitting longer dyads into smaller parts. [1-]
- ROLE
- is role assigned? [0 = false, 1 = true]
- SPEECHORD
- is speech in first or second half of dyad? [1|2]
- CONDITION
- (unknown - ask Robin) [A-H]
- SPEECH
- is speech allowed in this trial? [0 = false, 1 = true]
- GAZE
- is other eye gaze displayed in this trial? [0 = false, 1 = true]
- MOUSE
- is the other mouse displayed in this trial? [0 = false, 1 = true]
- TANGRAM
- Tangram used for this trial. [1-16]
- ORDERPART
- Position of this trial within this part of the dyad. [1-]
- ORDERDYAD
- Position of this trial within the whole dyad. [1-]
WARNING: at 18 Jan 08, we have discovered that not all of these fields get transferred by gdf2nxt.py into the NXT corpus resource, but we haven't traced through which do and do not. The metadata file does not include attribute definitions for the missing fields, either.
Fields (in left-right order) for experiment 3
- PAIR
- dyad id. [01-32]
- ROLE
- is role assigned? [0 = false, 1 = true]
- TANGRAM
- Tangram used for this trial. [1-16]
- ORDER
- Position of this trial within dyad. [1-16]
- INSTR
- is instruction "you can SEE your partner's mouse" [S] or "you mouse can be SEEN by your partner" [N]
- MOUSE_A
- is A's mouse displayed in this trial? [0 = false, 1 = true]
- MOUSE_B
- is B's mouse displayed in this trial? [0 = false, 1 = true]
- MSCOND
- 4 combinations of A and B's mouse display [A-D], defined as follows:
Condition A: -A, -B (symmetrical)
Condition B: +A, -B (asymmetrical)
Condition C: -A, +B (asymmetrical)
Condition D: +A, +B (symmetrical)
(+ means visible and - invisible)