Change windows on your browser to return to the program. |
Interact's View report form records results. However, the report may provide more information than you want. For example, you may be analyzing a series of events and want only a record of the deflections produced by the events.
The leftmost checkboxes on the Select options form allow you to focus on Interact results that are of interest to you. In all cases, checking a box causes Interact to print requested results into your browser's Java console. Text in the Java console can be selected and copied to a text-processing program.
Multiple checkboxes can be checked. For example, checking both the second and third boxes would present the deflection and actor's emotion profile for each event.
Clicking the Fundamentals, impressions checkbox causes Interact to print fundamentals and transient impressions for each event into your browser's Java console. The following shows the kind of text that is produced for an event.
Person 1[_,policeman],stop,Person 2[_,criminal]
Actor Fundamental: 0.65 1.48 -0.01. Behavior Fundamental: -0.04 1.48 0.27. Object person Fundamental: -1.93 -0.47 1.00.
Actor Transient inputs: 0.65 1.48 -0.01. Behavior Transient inputs: -0.04 1.48 0.27. Object person Transient inputs: -1.93 -0.47 1.00.
Actor Transient outcomes: 0.03 1.28 0.23. Behavior Transient outcomes: -0.15 1.35 0.30. Object person Transient outcomes: -1.17 -0.65 0.61.
Transients change from one event to the next. Fundamentals ordinarily stay the same in sequences of events, but they do change when reidentifications occur, or when Sentiment-Formation analyses are in progress.
Clicking the Deflections checkbox causes the following kind of result to print in the Java console when an event occurs.
Deflection: 1.27.
Clicking the third or fourth checkboxes on the left causes the following kind of result in the Java console.
Actor emotions 0.45 0.90 0.53
The numbers are the EPA profile for the emotion expected in the circumstances.
Clicking the fifth or sixth checkboxes on the left causes the following kind of result.
Actor's optimal act 0.44 1.04 -0.55
The numbers are the EPA profile for the behavior expected in the circumstances.
Who would do the action -0.27 1.24 0.43
The numbers are the EPA profile for the revised identity expected in the circumstances.
Actor characteristics 0.03 0.73 0.83
The numbers are the EPA profile for the attribute expected in the circumstances.
Clicking the Verbal events checkbox causes the following kind of output in the Java console for each event.
Person 1[_,adulterer],abuse,Person 2[_,adulteress]
Ordinarily you implement an event by clicking the event in the Analyze events form. This is tedious if you have dozens or hundreds of pre-defined events to get through.
Checking the Run-events-automatically box causes Interact to proceed to the next event and implement it after the last event is implemented. The processing is exactly the same as when you click events one after another.
You have to start an automatic process by clicking on the first event in the Analyze events form, in the usual way.
Interact slows down with batch runs of more than 30 events because it tries to archive all outcomes in reports. Unchecking the Record-all-events checkbox deals with this problem by causing Interact to stop writing reports.
Words are not reported in Interact analyses if the distance between their EPA profiles and the search profile exceeds a certain distance. The cut-off distance being used in Interact is displayed in the box labeled "Search cut-off distance". You can have Interact use a different cut-off by typing the value you want in this box and pressing Enter.
The cut-off value displayed in the box is used in searches throughout Interact.
You also can change the cut-off distance on the Find concepts form.
D. Heise's Understanding Events (New York: Cambridge University Press, 1979, pp. 61-63) proposed a general model of how fundamental sentiments might form from experienced transient impressions, and Interact implements some versions of that general model.
Sentiment formation can be turned on by entering a number greater than zero in the "Number of events to remember" box. Sentiment formation can be turned off by entering zero in the box.
If the From reidentifications box is checked then new sentiments are generated by averaging reidentifications in past events. If the From transients box is checked then new sentiments are generated by averaging impressions in past events.
A number of things happen when sentiment formation is operative in Interact.
A set of special concepts is added to the Interact identity list. Each consists of the word Mutator followed by a number representing the person who owns the concept and another number indicating who the concept identifies. For example, the identity Mutator_1_1 represents Person 1's view of self; Mutator_1_2 represents Person 1's view of Person 2, and Mutator_2_1 represents Person 2's view of Person 1. The EPA fundamental profiles for these identities will change as events occur. | |||||||||||
When an identity is selected on the Define situation form, the profile for that identity is attached to the appropriate Mutator identity and that Mutator identity is assigned to the person whose identity is being defined. For example, if the experiencer is Person 1 and "policeman" is selected to identify Person 2, then the EPA profile for policeman will be assigned to Mutator_1_2, and Mutator_1_2 will be shown as the identity of Person 2. | |||||||||||
Implementing an event
on the Analyze events
form changes the sentiments
attached to relevant Mutator identities for future events.
Specifically, if
we are analyzing dyadic events of Person 1 and Person 2 as viewed by
Person
1, and the number of events to remember has been set equal to R:
|
|||||||||||
The changing values of the fundamental sentiments can be examined on the View results form, or in the Java console if Fundamental, Impressions is checked. | |||||||||||
The actor's optimal behavior toward the object person is computed after all prior events, NOT including the current event (regardless of what shows on the Now-Next menu on the Analyze events form). Thereby you can compare optimal behaviors with actual behaviors. |
The model computes averages as follows:
While number of events is less than R, the EPA profile from each event is weighted by the reciprocal of the number of events that have occurred. | |
When number of events is greater than R, the EPA profile from each of the last R events is weighted 1/R, and profiles further back are weighted zero. |
Interactions involving more than two people: Averages are computed only over the events in which an individual actually participated. For example, if Person 3 did not participate in events five through ten, then those events would be ignored while computing the fundamental sentiment applying to Person 3.
Event definitions with the character "&" at the beginning define a set of events occurring simultaneously. See the example in Help on defining events.
Beginning a line with "$," stops effects of prior events and starts processing anew. See the example in Help on defining events.
Beginning a line with "#," stops effects of prior events and starts processing anew, with EPA fundamentals of all interactants re-set to 0, 0, 0.
Restarting can be combined with simultaneous events by listing the special characters in sequence. For example,
$,&>,Person 1[_,mother],talk to,Person 2[_,child]
would be an event following other events. The $ indicates that transients should be reset to zero; the &> indicates that this is the first of a set of events occurring simultaneously.