3.1 TARGET POPULATION AND SELECTION
3.2.4 Dependent variables
3.2.4.1 Parent behaviors
The following dependent variables were collected for parent behaviors: (a) contingent imitation, (b) linguistic mapping, (c) demands/questions and (c) imitation training.
Table 4. Parent dependent variables definitions and scoring
Dependent variables Definitions Scoring
Linguistic mapping The parent describes what the child is attending to or doing using simple language (e.g., ‘‘you built a tower) with or without sound effects.
Frequency count
Contingent imitation The parent imitates the child’s behaviors (i.e., actions with toys, gestures/body movements, and vocalizations) at the same time as the child engages in them, or within one to two seconds of the occurrence of the behavior.
Frequency count
Demands/Questions Parent asks the child a question or places a demand on the child.
Frequency count
Imitation training Combination of (up to three) parent model(s), prompt, and
reinforcement together or
combination of (up to three) parent model(s) followed by reinforcement (used when child imitates model spontaneously).
Frequency count
3.2.4.2 Child behaviors
Primary dependent variables collected for each child were prompted and spontaneous motor imitation. Social engagement was collected as a secondary dependent variable. See Table 5 and Appendix H for operational definitions and scoring information.
Table 5. Child dependent variables definitions and scoring
Dependent variables Definitions Scoring
Prompted imitation Child imitates the parent’s action (with or without a toy) with a physical/gestural prompt or verbal command within 10 seconds of the modeled behavior.
Rate per minute
Spontaneous imitation
Child imitates the parent’s action (with or without a toy) without a physical/gestural prompt or verbal command within 10 seconds of the modeled behavior.
Rate per minute
Social-engagement Combined total of five behaviors: 1) social gaze
2) mutual gaze
3) responding to joint attention, 4) initiating joint attention and 5) initiating behavior requests.
Total duration in seconds
3.2.4.3 Parent intervention fidelity
Intervention fidelity data were collected during the 10-minute parent-child play sessions that occurred in baseline, at the end of training sessions during the parent training condition, and at follow-up (see Appendix I for example of intervention fidelity calculation). Parents were taught specific strategies in each module that built upon one another to form a comprehensive
intervention program; therefore, two measures of intervention fidelity were calculated for each 10-minute parent-child play session: (1) module fidelity and (2) overall fidelity. Module fidelity was the fidelity rating for the techniques in the individual modules. Overall fidelity was an average of the fidelity rating for all four modules.
Parents were rated on a 0 to 5 scale for each module with a “0” for parent never uses RIT techniques and misses all opportunities and a “5” for parent uses RIT techniques throughout. Although the range was the same for each module (i.e., 0 to 5), calculation for each module was contingent on the type of techniques being taught to the parent.
The scale for module 1 (introduction to imitation training) included seven fidelity items (i.e., reduces physical area/creates defined space, de-clutters room, removes of distractions, identifies 4-6 sets of toys and places them on the floor, selects appropriate toys, removes overly absorbing toys). Each fidelity item included a detailed description of what signified a 0 – 5 rating. For example, in the fidelity item “reduces physical area/creates a defined space for intervention,” a “0” signified that an area has not been defined and the child may leave the play area more than 8 times during the session, engaging in challenging behaviors throughout and a “5” signified that a designated space is created by the parent using physical barriers such as furniture or a section of the room and the child never leaves the area.
The scale for module 2 (imitate your child) used a 30-second partial interval recording system to rate seven indicators that the technique “imitate your child” was performed correctly. Each interval for each indicator was coded as a “1” for the occurrence of the behavior, a “0” for the non-occurrence of the behavior, or a shaded cell for no opportunity to display the behavior. A sum was calculated for each indicator, converted into a percentage, and then converted into a 0 to 5 rating for each indicator based on a table that associated percentages with specific ratings. The seven ratings were then averaged to obtain an intervention fidelity rating session score for the targeted technique.
The mean score for module 3 (describing your child’s play) was calculated in the same manner as module 2, except only six techniques were coded. The scale for module 4 (imitation training) included one fidelity item describing how well and often the parent used imitation training techniques (i.e., 0 = never used the techniques, 5 = used a combination of the techniques five times across 3-one minute intervals). This variation in data collection across the four
fourth modules were better reflected through the use of a rating scale, whereas those used in the second and third module were evaluated more appropriately with an interval recording system that was then converted into a point on the 5 point rating scale. For example, module 1 taught the parent to set up the environment. The parent could only complete this task one time; therefore, it would not have been accurate to measure this behavior using a partial interval system.
Conversely, module 2 taught the parent to imitate the child; therefore, the parent had the opportunity to engage in this behavior at least one time every 30-seconds. An example of calculations for the fidelity rating is shown in Appendix I.