Download
1 / 56
590 likes | 908 Views
BEHV 415: Research I. 17 November 06. Agenda. Goals of this course Students will be able to provide the definition of a single subject design Students will be able to define each of the experimental designs reviewed in the class text
E N D
BEHV 415: Research I 17 November 06
Agenda • Goals of this course • Students will be able to provide the definition of a single subject design • Students will be able to define each of the experimental designs reviewed in the class text • Students will be able to look at graphs from behavior analytic journals and identify the experimental design • Given a research scenario, students will be able to identify the appropriate experimental design and discuss the advantages and disadvantages of using the design • Students will be able to define all of the terms from Behaviorspeak • Students will learn how to interpret quantitative data and create visual displays of quantitative data • Students will be able to define and determine when to use all of the direct and indirect data collection methods • Exam feedback
Agenda • Student evaluations • Topics • Review of additional designs • Review of key terms • Focus on precision teaching • Focus on verbal behavior • Focus on data analysis
Single Subject Research Design Other Names for SSRD (Elaboration) • Single case experimental design • Time series design • Small-N design • Single system designs • Within-subject comparison • Idiographic research • N of one trial
Single Subject Research Design • SSRD involves studying a single individual or system by taking repeated measurements of 1 or more dependent variables and systematically applying & sometimes, withdrawing or varying the independent variable. Ottenbacher, 1986; Bloom & Fischer, 1982)
Reasons to Use SSRD • Demonstrates individual differences • Difficulty obtaining adequate power (N) • Difficulty obtaining homogeneous group • SSRD is relatively easy to do • Helps validate practice • Demonstrate treatment effectiveness • Great pilot
Overview of the Study • To Help Or Not To Help • How can this study be transformed into an alternating treatments design? • Why are baseline data unnecessary? • Group I (summarize) • Group II (alternating treatment design) • Group III (Data collection and display) • Group IV (Linkages to terms and concepts) • Group VI (Relevant application of the design)
Advantages and Disadvantages • Advantage • Determine the effects of multiple interventions within a short period of time • Baseline data do not need to be stable before intervention (since you are comparing two interventions) • Disadvantage • Multiple treatment interference • Problem with reversibility
Adapted Alternating Treatment • Each intervention is applied to different behaviors that are considered to be of equal response difficulty but functionally independent
Stimulus Class • A set of stimuli, all of which have some common property • Stimuli that vary across physical dimensions but have a common effect on behavior
What is Abstract Art? • Art that departs significantly from natural appearances. Forms are modified or changed to varying degrees in order to emphasize certain qualities or content. Recognizable references to original appearances may be slight. The term is also used to describe art that is nonrepresentational.
Rule-Governed Behavior • The effects of instructions, advice, maxims, and laws on the listener's behavior. In this view, rules are seen as complex discriminative stimuli and the principles that govern stimulus control regulate the behavior of the listener
Rule-Governed Behavior • Behavior either verbal or nonverbal, under the control of verbal antecedents. In some usages, any verbal antecedent qualifies as a rule (as when one is told to do or say something). In others, rules are only those verbal antecedents that specify contingencies (as when one is told what will happen if one does or says something); such rules may alter the functions of other stimuli. Some rules are self-produced; the most effective verbal antecedents are those one generates oneself. Whether rule-following occurs in the presence of a rule is often ambiguous (one may repeat a rule to oneself at the time of following it); for that reason, rules do not necessarily qualify as discriminative stimuli even though they function as verbal antecedents
Stimulus Class Stella Moore
What is surrealism? • European literary and artistic movement that uses illogical, dreamlike images and events to suggest the unconscious.
Surrealism Dali Chagall
Stimulus Control • A phenomenon described by a very high probability of behaviors occurring (or not occurring) in the presence of particular antecedent stimuli, and a very low (or higher) probability of occurring in their absence
Changing Criterion Design • The changing criterion design involves the evaluation of the effects of a treatment on the gradual, systematic increase or decrease of a single target behavior • This accomplished by carefully changing, in a stepwise fashion, the criterion levels necessary to meet contingencies to increase or decrease behavior
Changing Criterion Design • The effect of intervention is demonstrated when the target behavior changes to the predetermined criterion levels specified by the experimenter
Changing Criterion Design • In the changing criterion design the level of responding of each phase (subphase) actually serves as a baseline for the subsequent phase
Precision Teaching Precision Teaching boils down to "basing educational decisions on changes in continuous self-monitored performance frequencies displayed on 'standard celeration charts'" (Lindsley, 1992a, p. 51). As such, it does not prescribe what should be taught or even how to teach it: "Precision teaching is not so much a method of instruction as it is a precise and systematic method of evaluating instructional tactics and curricula" (West & Young, 1992, p. 114).
Precision Teaching Focus on Directly Observable Behavior Frequency as a Measure of Performance The Standard Celeration Chart The Learner Knows Best
Precision Teaching Focus on Directly Observable Behavior To avoid ambiguity, Precision Teachers attempt to translate learning tasks into concrete, directly observable behaviors that can be counted and recorded. McGreevy (1983) recommends that the task be a physical movement, something the person is doing. For example, the physical acts involved in "putting on a sweater" could be observed and counted. Another alternative is to count a tangible product of something that has been done, such as "sweater on body."
Precision Teaching Frequency as a Measure of Performance In Precision Teaching, a behavior frequency is "the average number of responses during each minute of the assessment period" (White, 1986, p. 523). It is specified as counts per minute. Precision Teachers have written extensively about the advantages of frequency data over traditional measures in education such as percent correct (e.g., Binder, 1996; Lindsley, 1991; West & Young, 1992). Two of them will be noted here. First, frequency data are ultimately more useful. Fluent (i.e., accurate and high frequent) performance is retained longer, endures better during long time-on-task periods, is less likely to be affected by distracting conditions, and is more likely to be applied, adapted or combined in new learning situations, even in the absence of instruction (Binder, 1996; see also West & Young, 1992).
Precision Teaching The Standard Celeration Chart Lindsley recalls that "in desperation" he developed a standard chart. The y-axis was set up on a multiply scale to accommodate behavior frequencies ranging from 1 per day to 1,000 per minute; the x-axis was set up on an add scale to accommodate 140 successive calendar days, which is about the equivalent of one school semester. Thus, both frequent and infrequent behaviors could be plotted on the same graph and displayed for an entire semester. A goal of standardization was to save teachers time drawing and labeling the charts, and reading and interpreting the data. This latter point is especially important: "When teachers allow themselves the luxury of making a new chart for each behavior and learner, different pictures of progress are formed, the comparison of one program with another is difficult, and the evaluation of how well as program is working can be in error"
(White, 1986, p. 524). Figure 2 illustrates how the same data plotted on two different scales can suggest different impressions of a student's progress.
The multiply (logarithmic or ratio) scale of the y-axis on the Standard Celeration Chart offers other advantages. As West & Young (1982) note: "When the scores from repeated measures of performance are plotted on the more typical 'equal interval' or 'arithmetic' scale, learning (represented by a line or function which 'best fits' the data) is found to accelerate. In other words, a curve with an ever steeper slope is created. When data are plotted on the standard celeration chart, learning is generally represented by a straight or nearly straight line. The value of the slope of the line which best fits the distribution of values on the logarithmic scale is thought of as an 'index of learning' The steeper the slope, the faster the learning is; the flatter the slope, the slower the learning is." (p. 132).
The word celeration is the root word of acceleration and deceleration, the two ways in which behavior can change on the chart. When frequency doubles from one point in time to the next, it is said to be accelerating at "times 2," abbreviated as x2. When frequency is cut in half from one point in time to the next, it is said to be decelerating at "divided by 2," abbreviated as /2. Celeration is standardized in the sense that equal ratios of performance change appear as equal slopes on the chart, independent of the starting frequency. For example, a change from 1 to 2 (x2) would look similar to a change from 50 to 100 (x2). On an add scale, a change from 1 to 2 (+1 but x2) would be dwarfed by a smaller proportional change from 80 to 100 (+20 but x1.25). This is illustrated in Figure 3. Lindsley (1990a) suggests we abandon the terms "increase" and "decrease" altogether when describing behavior change. He claims that a major discovery of Precision Teaching is that all behavior "multiplies" or "divides" and that we best think in these terms.
Precision Teaching The Learner Knows Best A general rule of thumb in Precision Teaching is that if a student is progressing according to plan, then the program is appropriate for that student; otherwise, there is a flaw in the program and it needs to be changed in some way. In other words, the student's performance determines the "right" teaching strategy.
Students are active participants in Precision Teaching. Typically, students keep count of their movements on a daily basis and record the counts on a Standard Celeration Chart. This enables students to "see" their learning. The visual display reveals ongoing performance in relation to the goal, and is the basis on which the teacher and student decide what to do next. Lindsley (1990a, p.11) notes five advantages of students taking this active role: It costs less than teacher or observer recording. It produced records as reliable and much more valid than other recording. The effects produced were usually larger than teacher managed effects. It developed a trust of the learner in contrast to the erosion of trust produced by double checking of counts by teacher and observer. The learners developed higher order self-management skills to take with them in later life.
Single Subject Research Design:Analysis • Visual analysis most common • Assess trends & levels between adjacent phases • Level – refers to change in value or magnitude of dependent variable after intervention • Trend refers to change in direction • Described as accelerating, decelerating, stable or variable • Another technique Split middle Difference • Uses celeration line to which statistical significance can be explored • Also 2 standard deviation method
How to find the Median Value It's the middle number in a sorted list. To find the Median, place the numbers you are given in value order and find the middle number. Look at these numbers: 3, 13, 7, 5, 21, 23, 39, 23, 40, 23, 14, 12, 56, 23, 29 If we put those numbers in order we have:3, 5, 7, 12, 13, 14, 21, 23, 23, 23, 23, 29, 39, 40, 56 There are fifteen numbers. Our middle number will be the eighth number: 3, 5, 7, 12, 13, 14, 21, 23, 23, 23, 23, 29, 39, 40, 56 The median value of this set of numbers is 23. (Note that it didn't matter if we had some numbers the same in the list)
BUT, if there are an even amount of numbers things are slightly different.In that case we need to find the middle pair of numbers, and then find the the value that would be half way between them. This is easily done by adding them together and dividing by two. An example will help: 3, 13, 7, 5, 21, 23, 23, 40, 23, 14, 12, 56, 23, 29 If we put those numbers in order we have:3, 5, 7, 12, 13, 14, 21, 23, 23, 23, 23, 29, 40, 56 There are now fourteen numbers and so we don't have just one middle number, we have a pair of middle numbers: 3, 5, 7, 12, 13, 14, 21, 23, 23, 23, 23, 29, 40, 56 In this example the middle numbers are 21 and 23. To find the value half-way between them, add them together and divide by 2: 21 + 23 = 4444 ÷ 2 = 22 And, so, the Median in this example is 22.
How to Find the Mode or Modal Value The mode is simply the number which appears the most. To find the mode or modal value requires you to put the numbers you are given in order. Look at these numbers: 3, 7, 5, 13, 20, 23, 39, 23, 40, 23, 14, 12, 56, 23, 29 In order these numbers are: 3, 5, 7, 12, 13, 14, 20, 23, 23, 23, 23, 29, 39, 40, 56 This makes it easy to see which numbers appear the most. In this case the mode or modal value is 23.
Standard Deviation and Variance Deviation just means how far from the normal Standard Deviation The Standard Deviation (σ) is a measure of how spread out numbers are. The formula is easy: it is the square root of the Variance. So now you ask, "What is the Variance?" Variance The Variance (which is the square of the standard deviation, ie: σ2) is defined as: The average of the squared differences from the Mean. In other words, follow these steps: 1. Work out the Mean (the simple average of the numbers)2. Now, for each number subtract the Mean and then square the result (the squared difference). 3. Then work out the average of those squared differences. (Why Square?)
Example You and your friends have just measured the heights of your dogs (in millimeters): The heights (at the shoulders) are: 600mm, 470mm, 170mm, 430mm and 300mm. Find out the Mean, the Variance, and the Standard Deviation. Answer: so the average height is 394 mm. Let's plot this on the chart:
so the average height is 394 mm. Let's plot this on the chart: Now, we calculate each dogs difference from the Mean:
To calculate the Variance, take each difference, square it, and then average the result: So, the Variance is 21,704. And the Standard Deviation is just the square root of Variance, so: Standard Deviation: σ = √21,704 = 147
And the good thing about the Standard Deviation is that it is useful. Now we can show which heights are within one Standard Deviation (147mm) of the Mean: So, using the Standard Deviation we have a "standard" way of knowing what is normal, and what is extra large or extra small. Rottweillers are tall dogs. And Dachsunds are a bit short ... but don't tell them!
*Note: Why square ? Squaring each difference makes them all positive numbers (to avoid negatives reducing the Variance) And it also makes the bigger differences stand out. For example 1002=10,000 is a lot bigger than 502=2,500. But squaring them makes the final answer really big, and so un-squaring the Variance (by taking the square root) makes the Standard Deviation a much more useful number.
Analysis of Data • Level • Possible changes that would be viewed vertically. A jump in the data path, either upward or downward • Calculate performance • Mean • Median • Mode • Range • Create level lines (p. 207 handout)
