Data Collection & Analysis Presented by David Martin, Behavior Consultant, CMCSS Schedule 8:00 – 9:30 9:30 – 9:50 9:50 – 11:30 11:30 – 12:45 12:30 – 1:30 1:30 – 1:45 1:45 – 3:00 Work Break Work Lunch Work Break Work Materials Thank you to the author of, Behavioral Assessment and Program Evaluation, 2007 Dr. Jose Martinez-Diaz The majority of material in the presentation was taken from his book. Dr. Patti Wilson Some of the data collection forms came from her collection. Goals of the In-Service Review Target Behaviors Review Response Classes Discuss Fundamental Properties and Dimensional Quantities of behavior Data Collection and Display Discuss Data and Response Measures Continuous and Discontinuous Measures Selecting and Using Response Measures Activity - Pretest Which are examples of target behaviors? 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. Aggression Biting others Spitting on others Non-Compliance Following teacher directions Low Self-Esteem Social Skills Thinking Off-Task On-Task Showering Place your answers aside and we’ll return to them later. Do We Have To? Really? Goals are data-driven Data tells us if what we are doing is working Because everyone likes to see an effect – especially students Data is the key to perfecting a plan Data provides objectivity vs. subjectivity Yes. We Have to. BUT KEEP IT SIMPLE! Mechanisms convenient should be simple and Rate Use of a timer (latency, duration, etc.) + or – point rankings (0, 1, or 2) Establish a baseline Set a timeframe for the intervention(s) and an attainable goal Revise goal/interventions as necessary Identifying Target Behavior A target behavior is any defined, observable, and measurable behavior which is the focus of analysis and intervention. Identifying Target Behavior The behavior of an organism is that portion of the organism’s interaction with the environment that is characterized by detectable displacements in space through time of some part of the organism and that results in a measurable change in at least one aspect of the environment. Johnston & Pennypacker, 1980 Identifying Target Behavior Behavior is the movement of an organism or its parts in a frame of reference provided by the organism or various external objects or fields. Skinner, 1938 Identifying Target Behavior The interaction of the muscles and glands of a live organism and the environment. Martinez - Diaz Identifying Target Behavior What 1. 2. 3. 4. do they all have in common? Observable Measurable Interaction with the environment Change in the environment Identifying Target Behavior Examples of Non-Behavior Anorexia Depression Positive Attitude Repressed Feelings Poor Impulse Control Non-Compliance Lazy Defiant Disrespectful Identifying Target Behavior The Dead Man’s Rule If a dead man can do it, it is not behavior! Identifying Target Behavior A private event is something that we cannot see or measure – it is not behavior. Examples: Thinking is a private event Feeling is a private event Some argue that reading is a private event Activity - Pretest Which 1. 2. 3. 4. 5. 6. 7. 8. 9. are examples of target behavior? Aggression Biting others Spitting on others Non-Compliance Following teacher directions Low Self-Esteem Social Skills Thinking Showering Behavior: 2 3 5 11 Activity Get with five people close to you. Go over each non-behavior on the pretest. Identify why it is not considered behavior. Change each non-behavior item into observable, measureable language that could be used to develop a target behavior. Be ready to report out. Things I Need to Do… Reflect on a student you have had, or will have that emitted behavioral problems. If you identified a non-behavior as the problem, take a minute to change that language into target behavior language. Write the new language on your Things I Need to Do List. Options: You may want to consider a referral you’ve written, a behavior plan, or even a note or phone call to a parent. Identifying Response A response is a SPECIFIC instance of behavior However, behavior is a collective term. We don’t say: “Three yelling behaviors” or “Four Face slapping behaviors” Identifying Response Response Cycle A response cycle refers to the beginning, middle and end of a response. Examples: Holding one’s breath under water Face-slapping behavior Exiting a computer game Getting paper out for a test Identifying Response Topographical Response Class A topographical response class refers to the physical nature of responses. That is: the exact shape of the response, the appearance of the response, the force of the response, and the actual movements involved. Identifying Response Topographical Response Class So a topographical response class is a group of two or more responses that share a common form. Examples: Pointing Hand to Face Movements Hair Pulling Identifying Response Functional Response Class A functional response class is a group of two or more topographically different responses that all have the same effect on the environment, usually producing a specific class of reinforcers. Identifying Response Functional Response Class So a functional response class is a group of two or more responses that serve the same purpose. Examples: Attention-Getting Behaviors Food-Getting Behaviors Task-Escape Behaviors Activity In the next three minutes, on a scrap piece of paper, list as many task-escape behaviors as you can. These should be behaviors you have observed (or emitted). Things I Need to Do… You have just created a functional response class for task-escape. When you identify target behavior, you need to do this for whatever the target behavior might be. So, if a child wants attention, you identify all the behaviors in an attention seeking functional response class that this child emits. On your list, write the name of a student for whom an FBA/BIP is scheduled. Write the name of the student and a reminder to identify the functional response class (after function has been identified) to be included in the FBA/BIP. Fundamental Properties of Behavior A property is a fundamental quality of a phenomenon Dimensional Quantities of Behavior A dimensional quantity is a quantifiable aspect of a property. Fundamental Properties of Behavior A response occurs in time. More precisely, a response occurs at a certain point in time in relation to a preceding environmental event. So temporal locus (the point where a response is situated in time) is a fundamental property. Dimensional Quantities of Behavior The accompanying dimensional quantity of temporal locus is latency, which is the amount of time between a stimulus (asking someone to do something) and the response (starting to do it). We measure latency. Fundamental Properties of Behavior A second fundamental property of a single response is derived from the fact that a response also occupies time, thus the property of temporal extent (how much time the response takes). Dimensional Quantities of Behavior The accompanying dimensional quantity of temporal locus is duration, or the amount of time from the beginning to the end of a response cycle. We measure duration. Fundamental Properties of Behavior A third fundamental property of a single response is repeatability through time. It refers to the fact that a response can reoccur. Dimensional Quantities of Behavior Countability is the dimensional quantity associated with repeatability, which is measured as the number of responses, or number of cycles of the response. We measure frequency. Fundamental Properties of Behavior The fundamental property of a response class is the combination of repeatability and temporal locus. From this combination, we get more dimensional quantities. Dimensional Quantities of Behavior Inter-response time (IRT) refers to the time between two successive responses; usually the time elapsed between the end of one response cycle and beginning of the next response cycle. We measure IRT. Dimensional Quantities of Behavior Rate of responding is the ratio of the number of responses over some period of time. Many of us confuse rate and frequency. They are NOT the same thing. We measure rate. Dimensional Quantities of Behavior Rate is the fundamental datum in the study of behavior. Its unit of measurement is cycles per unit of time. So…three cyclers in one hour is: 1 per 20 minutes. ACTIVE STUDENT RESPONDING Active Student Responding is a researchbased strategy to keep students engaged in the lesson. Enjoy the Activity! ACTIVE STUDENT RESPONDING A single response can reoccur, thus the fundamental property of: 1. 2. 3. Repeatability Temporal Extent Temporal Locus ACTIVE STUDENT RESPONDING A single response can reoccur, thus the fundamental property of: 1. 2. 3. Repeatability Temporal Extent Temporal Locus ACTIVE STUDENT RESPONDING A single response occurs at a point in time, thus the fundamental property of: 1. 2. 3. Repeatability Temporal Extent Temporal Locus ACTIVE STUDENT RESPONDING A single response occurs at a point in time, thus the fundamental property of: 1. 2. 3. Repeatability Temporal Extent Temporal Locus ACTIVE STUDENT RESPONDING Which dimensional quantity is associated with temporal extent? 1. 2. 3. 4. Countability Duration Latency Intensity ACTIVE STUDENT RESPONDING Which dimensional quantity is associated with temporal extent? 1. 2. 3. 4. Countability Duration Latency Intensity ACTIVE STUDENT RESPONDING Which dimensional quantity is associated with repeatability? 1. 2. 3. 4. Countability Duration Latency Intensity ACTIVE STUDENT RESPONDING Which dimensional quantity is associated with repeatability? 1. 2. 3. 4. Countability Duration Latency Intensity ACTIVE STUDENT RESPONDING Which dimensional quantity is associated with temporal locus? 1. 2. 3. 4. Countability Duration Latency Intensity ACTIVE STUDENT RESPONDING Which dimensional quantity is associated with temporal locus? 1. 2. 3. 4. Countability Duration Latency Intensity ACTIVE STUDENT RESPONDING IRT and Rate share which fundamental properties? 1. 2. 3. Repeatability and Temporal Extent Repeatability and Temporal Locus Temporal Extent and Temporal Locus ACTIVE STUDENT RESPONDING IRT and Rate share which fundamental properties? 1. 2. 3. Repeatability and Temporal Extent Repeatability and Temporal Locus Temporal Extent and Temporal Locus ACTIVE STUDENT RESPONDING A student is given a worksheet with 100 math problems. He begins working on the first problem 10 seconds after he is given the worksheet. 1. Latency 2. Duration 3. Frequency (countability) 4. IRT 5. Rate ACTIVE STUDENT RESPONDING A student is given a worksheet with 100 math problems. He begins working on the first problem 10 seconds after he is given the worksheet. 1. Latency 2. Duration 3. Frequency (countability) 4. IRT 5. Rate ACTIVE STUDENT RESPONDING The student spends 1 minute and 20 seconds working on the first problem before he completes it. 1. 2. 3. 4. 5. Latency Duration Frequency (countability) IRT Rate ACTIVE STUDENT RESPONDING The student spends 1 minute and 20 seconds working on the first problem before he completes it. 1. 2. 3. 4. 5. Latency Duration Frequency (countability) IRT Rate ACTIVE STUDENT RESPONDING 1. 2. 3. 4. 5. After completing the first math problem, the student waits 5 seconds before he begins working on the second math problem. Latency Duration Frequency (countability) IRT Rate ACTIVE STUDENT RESPONDING 1. 2. 3. 4. 5. After completing the first math problem, the student waits 5 seconds before he begins working on the second math problem. Latency Duration Frequency (countability) IRT Rate ACTIVE STUDENT RESPONDING The student completes 30 math problems correctly. 1. 2. 3. 4. 5. Latency Duration Frequency (countability) IRT Rate ACTIVE STUDENT RESPONDING The student completes 30 math problems correctly. 1. 2. 3. 4. 5. Latency Duration Frequency (countability) IRT Rate ACTIVE STUDENT RESPONDING The teacher gave the student 1 hour to complete math problems. The student completed 30 problems in 1 hour. 1. 2. 3. 4. 5. Latency Duration Frequency (countability) IRT Rate ACTIVE STUDENT RESPONDING The teacher gave the student 1 hour to complete math problems. The student completed 30 problems in 1 hour. 1. 2. 3. 4. 5. Latency Duration Frequency (countability) IRT Rate Data Collection and Display There is more to data collection and display than lines and graphs. Data may be embedded in text. Example: Jim displayed an average of 6 acts of property destruction per 7-hour school day during the 10 day baseline period. The range was between 1and 18 per day. Data Collection and Display There is more to data collection and display than lines and graphs. Data may be displayed in a summary table as is the attendance data below. Day Absent Present M T W Th F X X X X Tardy Excuse d X Data Collection and Display In behavior analysis, bar graphs and line graphs are by far the most common. We’ll focus on those. Data Collection and Display Bar Graphs, or Histograms, are used for: Summarizing and comparing different groups Summarizing and comparing different conditions Summarizing and comparing sets of data Data Collection and Display Bar Graphs, or Histograms Bar graphs compare sets of data which are not related to one another by a common underlying dimension (by which the horizontal axis can be scaled). 6 4 2010 2 2011 0 2012 Math Reading Writing Science Data Collection and Display Equal Interval, or Line Graphs, are used for: Tracking student progress Evaluating intervention effectiveness Used to determine function of behavior Series 3 6 4 2 0 8:00 9:00 10:00 11:00 Data Collection and Display Equal Interval, or Line Graphs The absolute minimum number of data points to be used for analysis is three. Seven is a better number. Series 2 5 4 3 2 1 0 8:00 9:00 10:00 11:00 12:00 1:00 Data Collection and Display Line Graphs Some notes: When you are looking for significant change on a line graph, the rule of thumb is this – Doubling or halving the rate is significant. Smaller changes may be important, but not significant. Data Collection and Display Line Graphs Some notes: When using visual analysis (looking at the data represented on a graph) the key points to look at are: Number of data points Trend Variability Data Collection and Display Line Graphs – Trend Trend refers to the overall direction taken by the data path. Trends are described in terms of their direction (increasing, decreasing or zero trend “level”). Data Collection and Display Line Graphs– Increasing Trend 7 6 5 4 3 2 1 0 8:00 9:00 10:00 11:00 Data Collection and Display Line Graphs– Decreasing Trend 7 6 5 4 3 2 1 0 8:00 9:00 10:00 1:00 Data Collection and Display Line Line Graphs– Decreasing Trend w/ Trend 7 6 5 4 3 2 1 0 8:00 9:00 10:00 1:00 Data Collection and Display Line Graphs– Variability Variability is another way of identifying the stability of the data. Generally, the more stable, the more you can trust the data. The more variable, the more time needed to get more stability in the data. Data Collection and Display Line 7 Graphs– High Variability 6 5 4 3 2 1 0 8:00 9:00 10:00 1:00 12:00 1:00 Data Collection and Display Line Graphs– Low Variability 6 4 2 0 8:00 9:00 10:00 11:00 1:00 1:00 Data Collection and Display Scatter Plot The use of scatter plots is still rare (we tend to use line graphs). However, scatter plots are VERY helpful as we try to identify the time target behaviors are emitted, with an easy to see visual display of the patterns of responding. Data Collection and Display Scatter Plot M 8 9 10 11 12 1 2 T W Th F Data Collection and Display Scatter M 8 9 10 eat 12 1 2 T Plot W Th F If the behavior is attention seeking (functional response class includes: calling out, getting out of seat, approaching teacher, crying, repeated questioning – What COULD we conclude about the behavior? Choose a Data Collection Method Continuous (direct observation) The show me method Event Recording Frequency/Rate Latency Duration Discontinuous (indirect observation) The tell me method Percent occurrence/correct Discrete Categorization Partial Interval Recording Whole Interval Recording Momentary Time Sampling PLACHECK Examples of Continuous Measures Counting how many times someone bites another person Timing how long someone sucks one’s own thumb Timing how long it takes Josh to turn off the TV after he is asked to do so Event Recording Record time observation began Count the responses Record time observation ended Divide: Count/Time Unit Report as rate per (insert time unit) TARGET BEHAVIOR EVENT RECORDING RATE PER MINUTE Questions xxxxxxxxxx 2 per minute Self-Disclosures xxxxxxxx 1.6 per minute Positive Comments xxxx 1.25 per minute Duration Duration per occurrence Record each event in time and reset the clock Date: ____ Time: ____ To: ____ Setting: _________ Episode # Elapsed Time Per Episode ________ ________________________________ ________ ________________________________ ________ ________________________________ ________ ________________________________ Duration Total Duration Record the length of each occurrence and add all together Date: ____ Time: ____ To: ____ Setting: _______ Total Duration Per Session: ______________ Advantages and Disadvantages Advantages Continuous Measures of They are accurate Disadvantages Continuous Measures of They require one person to watch continuously over a set period of time They are hard for one person to use and teach at the same time They generally require longer periods of time to collect data Percent Occurrence/Correct Percent Occurrence Date: ____ Time: ____ Setting:___________ Opportunity #1: X Opportunity #2: ____ Opportunity #3: ____ Opportunity #4: X Percent Occurrence: 50% Percent Occurrence/Correct Percent Correct Date: ____ Time: ____ Setting: ____________ Trial 1: C or I Trial 3: C or I + + Trial 2: C or I Trial 4: C or I Percent Correct: 75% _ + A Note about Percentages The Dimensional Quantities of behavior include: - Duration - Latency - Countability/Frequency - Inter Response Time (time over cycles) - Rate (responses over time) - Celeration 1. Things I Need to Do… Reflect on a student you have had, or will have that emitted behavioral problems. Consider what you want to do to help with this behavior (increase a replacement behavior and/or decrease a target behavior). Write down the data collection that might help you accomplish these each of these goals. A Note about Percentages 2. Percentages DON’T access the Dimensional Quantities of behavior. We lose: - Number Correct - Number Incorrect - Number of Opportunities - Measure of Time A Note about Percentages 3. However, people are familiar with this data display and “think” they understand what it means. Discrete Categorization Codes: I – Independent; VP = Verbal Prompt; GP = Gestural Prompt; PP = Physical Prompt; MG = Manual Guidance Take the percent of opportunity to work each task, at each code level, and report as percent Goes to Sink I Rubs Hands w/soap GP Turns on faucet I Sets soap down MP Grabs towel I Wets hands I Rinses hands VP Dries hands VP Hangs Towel VP Picks up soap GP Turns faucet off MG Walks to towel I Partial Interval Recording Used to DECREASE behavior Record in time intervals of 10 to 20 seconds A response is recorded one time only if a BX occurs at any time during the interval 1 2 x Report 3 4 5 x x 6 7 8 9 x as a percent intervals 50% 10 x Whole Interval Recording Used to INCREASE Behavior Record in time intervals of 10 to 20 seconds A response is recorded one time only if a BX occurs during the ENTIRE interval 1 2 3 4 5 6 7 8 9 10 x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x Report x as a percent intervals 50% Momentary Time Sampling Record as occurring only if it occurs at the POINT in time in which an interval ends. It is recorded as a non-occurrence if it is not occurring at that precise moment in time. Set a timer for ten minutes. When it beeps, look up and record + (is occurring) or – (isn’t occurring. Report in percent time samples PLACHCECK A group of students is observed at the end of a timed interval. Set the timer for thirty minutes. Look up when it beeps and count the number of individuals engaged in the behavior. Compare with the total number of students Report as Percent of Students Engaging in BX. Advantages and Disadvantages Advantages of Discontinuous Measures They are easy to reproduce They are easy to chart Most people understand percentages A teacher can collect the data without an additional body Disadvantages Discontinuous Measures of They are less accurate ACTIVE STUDENT RESPONDING Active Student Responding is a researchbased strategy to keep students engaged in the lesson. Enjoy the Activity! ACTIVE STUDENT RESPONDING If I record the number of reports you submit each week, I am measuring: 1. 2. 3. 4. 5. Duration Latency Frequency Rate IRT ACTIVE STUDENT RESPONDING If I record the number of reports you submit each week, I am measuring: 1. 2. 3. 4. 5. Duration Latency Frequency Rate IRT ACTIVE STUDENT RESPONDING If I record the amount of time I spend on the treadmill each day, I am measuring: 1. 2. 3. 4. 5. Duration Latency Frequency Rate IRT ACTIVE STUDENT RESPONDING If I record the amount of time I spend on the treadmill each day, I am measuring: 1. 2. 3. 4. 5. Duration Latency Frequency Rate IRT ACTIVE STUDENT RESPONDING 1. 2. 3. 4. 5. I am taking data on a student with severe SIB behavior (eye-gouging). Due to the intensity of the behavior, I measure the time to the first response, and end my session. Duration Latency Frequency Rate IRT ACTIVE STUDENT RESPONDING 1. 2. 3. 4. 5. I am taking data on a student with severe SIB behavior (eye-gouging). Due to the intensity of the behavior, I measure the time to the first response, and end my session. Duration Latency Frequency Rate IRT ACTIVE STUDENT RESPONDING 1. 2. 3. 4. 5. To determine how often to provide reinforcement in a differential reinforcement schedule, I need to know the average time between responses. I need to measure: Duration Latency Frequency Rate IRT ACTIVE STUDENT RESPONDING 1. 2. 3. 4. 5. To determine how often to provide reinforcement in a differential reinforcement schedule, I need to know the average time between responses. I need to measure: Duration Latency Frequency Rate IRT ACTIVE STUDENT RESPONDING Recording whether a response occurs at all during a period of observation is an example of: 1. 2. 3. 4. Event Recording Duration Recording Whole Interval Recording Partial Interval Recording ACTIVE STUDENT RESPONDING Recording whether a response occurs at all during a period of observation is an example of: 1. 2. 3. 4. Event Recording Duration Recording Whole Interval Recording Partial Interval Recording ACTIVE STUDENT RESPONDING Dividing an observation period into segments and recording whether a response occurs for an entire segment is an example of: 1. 2. 3. 4. Event Recording Duration Recording Whole Interval Recording Partial Interval Recording ACTIVE STUDENT RESPONDING Dividing an observation period into segments and recording whether a response occurs for an entire segment is an example of: 1. 2. 3. 4. Event Recording Duration Recording Whole Interval Recording Partial Interval Recording Differential Reinforcement Typically when we use Differential Reinforcement Schedules, we take an average and set a goal for increasing, or decreasing the behavior just above or below the average and then adjust the target as mastery is met. Activity After considering how to use Differential Reinforcements, determine what you need to know (what data system to use) before delivering reinforcement. a. b. c. d. e. f. g. A student won’t stay in his seat A student won’t take out materials to begin working A student cries when given a task demand A student won’t stop using the computer when his time is up A student only works on a task for a few minutes prior to stopping A student won’t get in line when it’s time to leave the classroom. A student isn’t toilet trained Questions Contact information for David Martin: david.martin@cmcss.net Don’t use my office phone at Burt to leave a message. I can go for days and not be in the office. I RARELY miss checking my email.