1. What are the components of a graph?
2. Describe the components of a research design.
3. Give the two criteria of a functional relationship.
4. Describe the five different major research designs, and their pros and cons.
Why is it important to collect accurate data?
(Martin & Pear, 2015)
- Dr Lynn Caldwell, a behaviour analyst, was consulted by a mother whose 6-year-old son was “driving [her] up a wall”
- boy would slam the kitchen door closed every time he went out of the kitchen
- Dr Caldwell had the mother record ________ data: tallying each instance of the target behaviour on a sheet of paper on the refrigerator
- results: over a 3-day period, the total number of door slams was 123
- behavioural treatment program:
• mother was to provide verbal ________ each time the boy went through the door without slamming it
• if the boy slammed the door, however, he had to go back and wait for 3 minutes (____-___), then go through the door without slamming it
- results:
• after applying treatment for 3 days, the mother complained that the treatment was not working
• but the number of times the door was slammed had _______ to 87
• the mother was encouraged to continue, and the number of door slams quickly fell to about 5 per day
- conclusion: only accurately recorded data allow conclusions about the _____________ of behavioural treatment programs
Components of a line graph:
1. x-axis and y-axis lines (these may not be added automatically by Microsoft Excel)
2. ______ for x-axis and y-axis (time vs. target behaviour dimension) and units (e.g., days vs. %)
3. numbers on x-axis and y-axis
4. data points and lines connecting them
5. _____ line(s), if there are different phases (in Microsoft Excel, manually add a “line” shape)
6. phase label(s) (in Excel, manually add text boxes)
e.g., baseline, treatment, follow-up
Dos and Don’ts:
• Do include client name or initials, behaviours, or settings in research with multiple baselines (place in lower-right corner).
• Do use a data _______ (e.g., table) to go from the raw data in data sheets to a graph.
• Don’t put different target behaviours (or multiple measures of one target behaviour) on the same graph.
• Don’t use a ______ (automatically generated by spreadsheet programs) to indicate phases.
• Don’t use gridlines.
• Don’t add a title.
• Don’t use gray for axes and units, use black.
• Don’t connect baseline phase curve to treatment phase curve.
Basic components:
• ___________ variable (IV): the treatment applied to the client
e.g., applying punishment by scolding child
• _________ variable (DV): target behaviour measure
e.g., child raising voice to call for parent
• ___________ (or extraneous) variable: another factor that may affect behaviour that is not controlled for (avoid this)
e.g., when applying punishment to a child, a parent also pays attention to them
• __________ relationship: treatment procedure regularly causes a change in the target behaviour; requires two criteria:
1) IV-DV relationship: changing IV causes change in DV
2) replication: consistent pattern of results
Behaviour modification research typically uses single-subject research methods, and does not employ much ___________ analysis.
A-B design:
- has one ________ phase (A) and one treatment phase (B)
- in baseline phase, collect at least three data points that do not show a trend (which would indicate behaviour change) Note: some studies presented in this course violate this guideline!
- pros & cons:
advantage: satisfactory for self-management
not a true research design: lacks replication, so it cannot ascertain cause & effect; change may be due to confounding variable
e.g., nail biting reduced because fingers got slammed in a door--not because of treatment
A-B-A-B (or ________) design:
- has two baseline phases and two treatment phases
- examines same behaviour, in the same client, in the same setting
- can be varied to include more than one kind of treatment
e.g., A-B-A-C or A-B-C-A-C can compare B and C treatments
- pros & cons:
advantage: can establish cause and effect
potential problem: it may be unethical to ________ a beneficial intervention
e.g., morbidly obese person may gain weight back during second baseline
potential problem: behaviour may not revert back in second baseline
e.g., tennis player who learned a new, better way to serve may not go back to old way
Multiple-baseline designs:
- more than one A-B design is carried out
- characteristics:
• baselines may vary in length before treatment begins
• intervention phase is staggered across separate designs
• can provide evidence that treatment is _________
- multiple baseline across ________: apply treatment to several different people
e.g., nurses wearing protective gloves with HIV-positive patients
|
time |
|
|||
A-B condition 1 |
A |
B |
nurse 1 |
||
A-B condition 2 |
A |
B |
nurse 2 |
||
A-B condition 3 |
A |
B |
nurse 3 |
||
- multiple baseline across __________: apply treatment to several different target behaviours
e.g., proper pronunciation of “th,” and “z,” and “zh” sounds
|
time |
|
|||
A-B condition 1 |
A |
B |
pronounce “th” |
||
A-B condition 2 |
A |
B |
pronounce “z” |
||
A-B condition 3 |
A |
B |
pronounce “zh” |
||
- multiple baseline across ________: apply treatment in several different settings
e.g., treatment of stuttering at home, at work, in public
|
time |
|
|||
A-B condition 1 |
A |
B |
home |
||
A-B condition 2 |
A |
B |
work |
||
A-B condition 3 |
A |
B |
public |
||
- pros & cons:
advantage: has no reversals; thus, good for behaviours that do not ______
if behaviour change occurs only when the B phase starts in each condition, we can conclude it was the result of the intervention
potential problem: treatment may ______ across subjects, behaviours, or settings during supposed baseline phase
e.g., a man who believed he was a woman named Amanda also believed he was Leonardo da Vinci and Jesus in past lives; treatment for “Jesus” delusion also reduced delusion he was da Vinci (Lowe & Chadwick, 1990)
___________-treatments (or multielement) design:
- baseline and treatment (or two treatment) phases are applied in rapid succession
- extraneous factors that could affect the results (like time of day, therapist administering the treatment, location of treatment, etc.) can be counterbalanced
e.g., treatment might be given in the morning one day and in the afternoon the next day
- treatment effects shown by _____________: consistent vertical separation between treatment curves
e.g., standard tutoring vs. modified (peer) tutoring (Kohler & Greenwood, 1990)
- pros & cons:
advantage: extraneous variables will have less of an effect
advantage: evaluates effects of different treatments
potential problem: treatments may interact with each other
Changing-_________ design:
- criterion for successful treatment progressively changes (usually becoming more stringent over time)
- uses A-B design, but goal for target behaviour changes in the treatment phase
e.g., number of revolutions per minute pedaled on a stationary exercise bicycle by obese boys (De Luca & Holborn, 1992)
- pros & cons:
advantage: well-suited to behaviours that can be approximated gradually
potential problem: unsuitable for behaviours that may not change _________