Lesson 2 Data Collection
Contents
Programming decisions are made based upon data so the accurate collection of data is very important to any program. This lesson reviews how we typically take data on home programs but it's important to know that adjustments are made based on the individual needs of each child.
Uses of Data
All decisions made in an Applied Behavior Analysis program are data-driven and objective. It is important to keep data of targeted and mastered skills for a variety of reasons. First, it is the best way to accurately and effectively track progress. Second, we can build on and expand to new skills using the data we’ve collected on previously mastered skills. Lastly, constant collection of data ensures that we have a record of the skills we are working on.
Types of Data Collected
While the data is important, it must never interfere with our teaching. Data can be collected on as trial-by-trial data, where data is collected for every presentation of the Sd. Data can also be collected as single-probe data, where the data is collected on the first presentation of the day. Both single-probe and trial-by-trial data are typically collected with a plus (+) or independent (I) for a correct response and a minus (-) or prompted (P) for incorrect responses. Data can also be collected using tallies, which is mainly used when we are interested in determining the frequency of a behavior, i.e. mand data is collected as a tally of I and P over a time period (min) and then calculated as a ratio of independent and prompted mands per minute.
Probe, Teaching Trials, Mastery Criteria
Single-Probe: An easy and efficient way to collect data is through probes. The probe is collected the first time the target is presented that day and shows us whether or not the child can accurately and fluently provide the required response without prompts.
Teaching: We may occasionally also keep track of how often an item is taught throughout a session to ensure that the item is being targeted frequently enough
Mastery criteria: Typically based on the child’s learning history and often ranges between 2-4 consecutive independent responses or a percentage, typically 90%, but may vary.
Program Board
Program Board
For the “early learner,” all initial targets should be taught in the natural environment. This insures that the targets chosen are ones that the child finds relevant or important within a given toy or activity and increases the likelihood that he will “talk” while playing with the toy or engaging in an activity on his own. Some natural environment teaching will occur throughout the child’s daily activities. For example, he may request food items at lunch time or his blanket at bed time. In addition, language instruction can be “contrived” around specific toys or activities that the child finds interesting.
It is recommended to use a program board to document the different targets taught within the child’s activities. Write down all the requests the child makes. Transfers to other operants can be done anytime the child requests with no prompting within a given session but don’t move the target to a different verbal operant category until he is spontaneously requesting the item on the first attempt of the day. This will help insure that multiple transfers have been conducted before the target is in the new operant category. For example, if the child spontaneously (without any prompts) asks for the shovel, you might ask the child, “What’s this?” (R= shovel) while he is digging (tact). Later in the session, right before pouring (the reinforcer) ask again “What’s this?” as you hold up the shovel so you are gradually fading the tact response from the request. During the next session, you might “probe” the tact of the shovel by asking “What’s this?” when he doesn’t WANT the shovel.
Remember that the order in which you teach the different operants may differ depending on the specific skills of the child but starting with the request (mand) will help insure the child continues to WANT to play with the toy or engage in the activity. As you build the number of different operants taught within an activity, be sure to keep the ratio of requests very high compared to other operants. Avoid “teaching too much” and “playing too little” or the child may no longer be interested in playing with you!
Probe Data
Probe Data
- Data is an important part of running a home program but should never interfere with teaching. The data taken must be correct and valuable or it will just be a waste of valuable teaching time.
- Probing refers to “testing” the target the first time it is introduced during the day. The goal is to minimize the “teaching effects” of the previous day to determine if the child has maintained the correct response the following day before any teaching has occurred.
- Targets are probed and data is taken once as either “Independent” or “Prompted”.
- A target is considered “Independent” only if the child responds accurately, with no prompting, within 2 seconds.
- If the child does not respond within 2 seconds, the response is “Prompted” and the data is recorded as such.
- If the child responds incorrectly, the correct response is taught and the target is recorded as “Prompted.”
- Depending on the number of programs running and the number of current targets, probe data can be taken in different ways. If the child has many current targets, it may be best to probe a different verbal operant each day. For example, every Monday, probe tacts, Tues, probe receptives, Wed. probe intraverbals etc… If the child has fewer targets, they can all be probed at the beginning of the first teaching session of the day.
- An important consideration when determining how often to probe current targets is the learning history of the child. If it typically takes at least one week of solid “teaching” for a child to respond correctly with no prompting, taking probe data daily would not be necessary or advised.
- Some children can “handle” just probing the new targets on a very rich schedule of reinforcement or while engaged in a reinforcing activity. (So long as it’s not “thematic” in nature. We don’t want to teach the child to ask random questions!) Others will require a great deal of mixing in with mastered responses to keep them successful. If this is the case, as with teaching, keep the ratio of “mastered” to “acquisition” targets at 80% or higher.
- Once the probe is completed, review the results to determine what needs to be focused on during the teaching session.
- The number of Independent probes required for “Mastery” is also dependent on the learning history of the individual child. Some children will master a target within one teaching session and others will take a week. An “average” 3 independent “cold probes” but make sure this isn’t slowing you down. If you see the child is consistently responding independently by the second day of teaching and shows no signs of losing mastered skills, drop the probe requirement to 2. On the other hand, if you find the child is frequently, “losing” mastered targets, you may want to increase the number of “cold probes” required to determine mastery to 4.
- It’s always a good idea to have at least 2 different people taking probe data to be sure there are not “instructor variables” confounding the data.
Gabrielle Trapenberg,MA,ABA/BCBA and Tracy Vail.MS,CCC/SLP
| Attachment | Size |
|---|---|
| (2.02)+Data+Collection.doc | 33.5 KB |