A little more on data

This may seem a little redundant but it is crucial. We have already talked about data. For some people data is just something that has to be done to satisfy someone else who is paying the bills. It is sometimes no more than an afterthought or something to manipulate in order to either continue to do what they think is best or to continue to be paid. This may sound harsh, but unfortunately it is true. I have seen people do all of the above and heard people admit to doing these things. In some situations there has been national scandal resulting from the uncovering of fraudulent, lazy, or poorly designed data practices and reporting. Some times therapists/interventionists do what they believe to be the best thing for the individual. Without good appropriately interpreted data, it is difficult if not impossible to know what the best thing is for the individual or even if what is believed to be good is helpful.
Three of the most common honest problems with data are:
1. Lack of clarity about what data is being collected;
2. Additonal variables; and
3. Too complicated.

Note: Sometimes people cheat on their data because of one or more of the above problems. Assuring clarity, parsimony and that you know are really measuring what you want to measure helps to increase validity and is one essential component to improved outcomes.

Additional note: I’ve liked the word “parsimonious” from the moment I heard it and learned its meaning. It may be my rather dry sense of humor. Years ago in graduate school I learned that one of the four assumptions of science is that it is parsimonious. I found this rather ironic because it was explained to me that the meaning was that it was short, simple and to the point. The word “parsimonious” of course is not short or simple. According to Wikipedia, the free encyclopedia; “Parsimony is a “less is better” concept of frugality/economy/stinginess or caution in arriving at a hypothesis or course of action. The word derives from Middle English parcimony, from Latin parsimonia, from parsus, past participle of parcere; to spare”. And “Parsimony: The simplicity with which a theory explains phenomena”. From Practicing Educational Psychology; by Margaret M. Clifford (1981).

1. Problem one is usually resolved by a well written and well thought out Measurable Behavioral Objective, which is based on good evaluations/assessments (Including a good functional analysis of behavior whenever possible and appropriate). Please ask questions in the comments section if you have any.

2. Problem two is often helped by a good Measurable Behavioral Objective but unfortunately it is not quite as simple as that. I remember reading a paper that stated that the data demonstrated the effectiveness of an intervention and yet they compared samples from an urban area where services were available against samples from a rural area where the services were not available. There were so many possible additional variables or additional reasons why the children in the city progressed at a different rate than the children in the rural setting that the entire study lost validity and any real credibility. When you read the research design from a paper, you may note design flaws as large as or larger than the one mentioned. There are a few things that you can do to decrease tainted data resulting from variables which are unintentional and/or unaccounted for.
a. Separate data by: setting; therapists; times; and additional possible distractions. While your goal may be to increase a behavior across settings, therapists, and other environmental factors, and your overall data report may include everything, it is often a good idea to collect data separately for the purpose of analysis and intervention refinement.
b. Assure that intervention is provided consistently and as written.
c. Collect data in as unobtrusive manner as possible (this will be discussed in a little more detail under problem 4).

3. Problem three is helped by keeping it as simple as possible. (Keeping it simple can also help to keep it less obtrusive. Collecting data in public in a more obtrusive manner is not respectful of the individual and in any setting can be an additional variable affecting behavior.) A couple of ways to keep data collection simple and less obtrusive is through the use of a counter. You can even use two different counters in different pockets allowing you to collect data on a couple of different behaviors. Another option is the use of a stop watch, which can also be kept in a pocket. In some situations you do not have to collect data all the time. You can take samples; however, it is imperative that you sample across environmental variables. If you do take samples, it may add to validity to intermittently take data all the time to better assure accuracy and validity. Obviously these ideas will not work in every situation. Brainstorm how you can make it simpler. Discuss options and ask questions in this section (do not share confidential information).

Note: Make sure that you collect data in the same way that the objective is written. If you write an objective around increasing a behavior, do not record data around a different and possibly decreasing behavior. If you write the objective such that Johnny will increase the number of times he does something, do not record data in percentages etc.

Click here to continue with this presentation: Now try it out