Test 9 Flashcards
S why
- noise distracts drivers
- higher fatalities in bus accidents
- have used music in past for ID students on bus
S1 subjects/setting
6-8th grade students
max 44 kids
25min ride
S1 apparatus
Noise guard
Noise guard threshold
Light display
S1 Noise guard
- like a microphone set up on the ceiling to track noise levels
- noise above 500Hz would activate it (length)
- recorded the umber of events over the threshold (in 10ths of seconds)
- threshold 93dB
S1 Light display
during treatment panel of 10 red lights on win shield
-went on to provide feedback every 0.1s above threshold
S1 behaviours
Out of seat
Roughhousing
Coercive comments
S1 Out of seat
any posterioral disconnection with the seat
S1 Roughhousing
aggressive nonverbal interaction
S1 coercive comments
-by driver
threats, warnings, reminders calling a Childs name without a comment within 3s.
S1 observation system and recording
observer coded data and recorder in bus
- 15s observation intervals (allowing 5s to record)
- recorded instances of target Bx per interval
- tape of drivers comments and scored later
- 10BL in afternoon to control for reactivity
S1 reliability
- once during each condition
- intervals agreed/total
- Divers comments: smaller/larger
S1 BL
kids told there would be an observer
S1 treatment
lights, music, raffles
-firsst day experimenter explained:
Lights on when too loud
Music on first day all times, after only when noise was low
Raffle for hamburger coupons each day if cut off criterion was not exceeded (previous day)
-students could make music suggestions
Raffle for 10 movie passes after 5 consecutive days
If one student intentionally turned on lights they would be reported to the principle
Driver told to do the same
S1 treatment phaseout
last week burger coupon raffle eliminated
movie pass changed from 10kids from 5 days, to 2 passes daily
music still everyday
S1 setting the cutoff criterions
could exceed threshold for limited time before music and raffle lost
- 1dt day 26s
- 2nd 13s
- 3rd 5s
S1 design
ABABC reversal
S1 results
BL 624.1 noise outbursts per trip average duration 62.3s
T: drastically reduced
BL: increase
Treatment phase out: results maintained
out of seat Bx decreased, but gradual increase ov er treatment
Roughhousing - decrease
Driver coercion - decreased during treatment 17-2.2
S1 discussion
- end of trips when criterion was exceeded
- had an effect on other Bx as well (out of seat/R)
- coercions from driver also decreased
- was liked by students
S1 limitations
- Out of seat/roughhousing recovered throughout treatment (addressed in E2 by lowering noise levels)
- need raffles not everyone may have access to prizes
S2 subjects and setting
6-8th grade
afternoon
44 max kids
S2 apparatus
Same as Exp 1
S lights instead of 10
-90dB SPL*
S2 Bx
-out of seat and roughhousing were recorded in alternating 15s record intervals
noted if one or more students doing Bx
-driver comments same as E1
S2 BL
This experimental condition was conducted identically with that of Experiment 1. Treatment (lights and preferred music). The treatment procedure was introduced and con- ducted in the same manner described in Ex- periment 1 except that no raffles were available. In addition, the cutoff criterion was reduced more gradually in this experiment.
S2 Exp design
ABAB reversal
S2 results
noise outburst dropped
-still effective without raffles
S2 survey
kids and driver reported bus ride was better
S2 next
test the ages and populations
P Why
continuation of past research, see if attention was the IV increasing Bx
P Subjects
from inner-city school
3 kids
1K, 2 1st grade
all have low school attendance records
P observations
recorded morning and afternoon for attendance
observer would periodically enter rooms to check
P BL
took data for attendance
P principal praise
entered classroom each day and attend to student rearing presence in school
pat on back/few words of praise
duration around no longer than 60s
12th week
began delivering intermittent praise 2-3 times per week
P results
all participants increased attendance
during intermittent praise attendance decreased slighly
P2 subjects
3 5th grade boys
capable but academically non productive
P2 procedure/observations
-tutor procedure was implemented, visited daily for reading and math (word rec/addition worksheet)
-done outside of class, 1:1, little interaction
-recored, Tx only instructs and answers Q
around 13min
-second observer sat in occasionally
P2 IOA
38%? word reg 80-100% M 95
1/4 of math 87-100 M 98.8
P2 BL
Go through M/R tutoring, receive feedback on CR, no schedule attention for CR
P2 principle praise
principle attention contingent on increase in CR
- would tell kid he wants to see improvement end of BL
- if criterion met, brought to P and 30s-60s of praise. Same for math
P2 design
within and between subjects multiple baseline design
P2 results
Principle praise increased academic performance in both subjects
P3 subjects
2 3rd grade classes
36-38 kids each
most and least advanced classes
P3 procedure
6 math worksheet with 100Q
- one sheet presented each day (recycled order/prez)
- 27 sessions in experiment
Exp would enter class and hand out sheets. 5 mine timer set and begun. kids did as many as they could in timer period
scores posted in class so they could receive feedback
P3 reliability
approximately every day
exp would rescore 1-2 papers
agree/dis times 100
M98%
P3 BL
Papers administered to both classrooms
P3 Principal praise
list of students who improved and list of top highest scores
- principal came to class and asked students on the list to stand and praised them
- took less than 3min, done twice a week
- teachers did the same as BL
Procedure introduced in 2nd class 6 sessions later
P3 results
both classes showed increased scores considerably
P1 attendance B/A
50-80%
P2 rational
Most time nonproductive students get ignored and the bad ones get attention.
-didn’t involve classroom teacher to isolate variable but in real life should included teacher as well.
P3 design
multiple BL across classrooms
P3 conclusions
can be applied to Ind/grp and easily.
H purpose
find effective method to load horses using R+ not aversive procedures
-horse fighting loading can be dangerous
What is meant by the criterion of evaluation?
What will be the standard of proof that the independent variable actually influences behavior? Kazdin delineates between two possible criteria: experimental and therapeutic.
What is meant by experimental criteria?
The use of an experimental criterion signifies a reliable or veridical difference in behavior under experimental conditions. When using statistical procedures to evaluate an intervention, this criterion is intuitively labeled statistical significance.
What is meant by therapeutic criteria?
A therapeutic criterion signifies the clinical significance of any observed differences induced by the independent variable (Kazdin, 1999). This criterion is tantamount to asking, is the effect meaningful to the patient?
Applying these two criteria then, an intervention may reliably diminish a problematic behavior (experimental criterion), yet only to the degree that has very little benefit to the individual’s life (lack of therapeutic criterion). Ideally, the analyst will consider both the experimental (statistical) and therapeutic (clinical) significance of any selected criterion that is used to draw conclusions about a conducted intervention.
What is visual inspection?
Visual inspection is looking at the graphed data to determine next steps?
What is statistical analysis?
collecting and scrutinizing every data sample in a set of items from which samples can be drawn
What is continuous time-series?
Continuous time series involves uninterrupted collection of observations.
What is discrete time series
A discrete series is one in which the observations are made at equal intervals in time, with the intervals between observations remaining constant throughout the observation period.
What is “shock”?
unique influences that drive data up or down its mean level at the time.
What is autocorrelation?
When observations of a phenomenon are collected repeatedly over time, subsequent observations tend to be more related than observations more temporally distant
What is non independence?
when observations are not independent, observations close together in time become predictive of one another. This nonindependence creates data structures that when described linearly, produces residuals, or error terms, that are correlated.
What is lag 1?
Lag 1 correlation represents the degree to which the current observation can be predicted from the observation prior. It would be calculated by correlating the second observation with the first observation, the third with the second, and so on throughout the series.
H purpose
The purpose of the present study was to develop and validate a program based on principles of positive reinforce- ment, as well as to create a fast and safe way to get horses to ‘‘self-load.’
H define self loading
Self-loading was defined as the trainer being able to walk the horse to the trailer, toss the lead rope over its back, say ‘‘step up,’’ have the horse walk into the trailer, and stand still long enough for the trainer to latch the butt chain and close the trailer door.
H what was the target used for
trained to touch with nose, this would prompt behaviour
H horses
5
took up to 3hrs to load
had been forced in the trailer in the past
no trailer sessions 6 months prior
H settings and material
in a field
trained ind
2 trainers each session (one data/R+ then one handled)
H trailer
white, two horse, butt chain, both doors open
H target
red cloth pot holder
H sessions
not done within 2hrs of feeding
lasted around 15min
each horse had one per day
H measurement
1-approached trailer within3m of entrance 2-approached trailer to entrance 3-approached until head and neck in trailer 4-A front legs 5-A half body 6-3 legs 7-4 legs less than 5s 8-all the way in and butt chain latched
H DV
Freezing Head toss Rearing Standing Turning Getting into the trailer Loading
H freezing
planted feet and refused to move
H head toss
horse jerked head away from trailer
H Rearing
front feet leave ground at same time
H Standing
went in front of T but not go in when prompted to
H turning
positioning body so no longer straight
H getting into the trailer
all four legs but didn’t satay long enjoy for butt chain to latch
H loading
all four legs in and butt chain latched, closed door
H prompt
VP step up
-new lead if walked away and came back.
H data
taken on how long it took to respond when presented with the target
within 5s
after 5s
nr
H reliability
live records vs videotape
25% of sessions
100%
H BL
once a day for 5m
- H lead to trailer and prompted to step up every 5s
- if moved head away - new lead
- froze - circle
H target training
clicker paired with food
training began on 5th session for all
prez with target, if touched clicker activated and gave food
target close to start, then increase/randomize distances
80% within 5s over 2 sessions move up a level
-word touch added
90% over 5 sessions moved to trailer training
H trailer training
led to trailer
-prompted to touch target-2nd gave R+
-Then prompted out of the trailer after trial
target moved
-if IR horse led out and start a new trial
-when consistent, moved target
-once always in trailer extension introduced , made trailer longer
-when in trailer for 10s butt chain lacked , but door left open, after 10-15 CR door closed
-next extension removed
-next loading on right side
-switching sides
-new lead
mastered after this
H Fancy
required additional training
2nd extension
limited hold implemented (to reduce prompts)
companion animal
H design
multiple baseline design across horses
H results
all horses learned to touch the target in first session
all showed a reduction in unwanted Bx
all learned to load
H S
all trained without punishment
other Bx effected as well
H discussion
1-target training stimulus control
2- not ind process used similar procedure for 4/5
3- small changes disrupted loading (switch L-R) test for gen
4- compliance improved outside of session
what is lag 10
looking at the autocorrection over the past 10 data points.
What is serial dependency
serial dependence if the value at some time t in the series is statistically dependent on the value at another times.
What are the 3 limitations of t test and F test
- these tests are robust in terms of normality
- assume obs in each phases are independent (no AC)
- insensitive to trends that occur within a phase
What is randomization
it is assumed that exp conditions are randomly assigned
What is alternation randomization
randomly assigns the intervention to given observations using predetermined limitations
What is phase randomization
- intervention is optimally introduced over the course of several measurement occasions occurring in a predetermined order.
- commonly used when interventions are predicted to have a carry over effect
What is meant by abrupt or gradual onset
quick or slow. Matched with duration
What is meant by permanent or temporary duration
Matched with onset type.
What is change in level
a comparison between the ending value of phase 1 and the starting value phase 2
what is change in slope
a linear rate of change.
What is a celeration line
The Overall celeration describes all of the frequencies charted, regardless of cycles, trends, periods, or event manipulations. An Overall celeration runs as a single line of best fit through all of the frequencies.
What is meant by replication is at the heart of any science
replication serves at least 2 purposes:
- Reliability of findings – can others get similar results with same procedures
- Generality of findings- do you get similar results under different conditions
- Interrelated, every time you get generality you also get reliability
What is direct replication
-Replication of a given experiment by the same investigator
Repetition of the experiment on the same subject – across behaviors for example
Repetition on different subjects – same procedures applied to different subjects
How is replication related to generality
.
What is simultaneous replication
Useful when intervening with a coexisting group
Number of subjects treated simultaneously in a particular experimental design, plot individual data
What are the guidelines for direct replication
- Therapists and settings constant
- Behaviour topographically similar across subjects
- Client background variables closely matched (homogeneity)
- treatment should be uniform (unless failure)
- 1 ex, 3 rep are sufficient
- still doesn’t provide broad generality but enough to work with
What is systematic replication
Replicate findings from a direct replication series varying settings, behavior change agents, behavior disorders, or any combination thereof.
Provides further information on generality of findings across clients because new clients are usually included
Why are failures important
- determine cause of intersubject variability
- continued research
What are the guidelines for systematic replication
- should start with sound direct replication
- change 1-2 variables
- is a search for failures, if none then continue
- When is it over, never over
What is clinical replication
Replication in which a treatment package containing two or more distinct procedures is applied to a succession of clients with multiple behaviors, i.e. “parenting package”
Field Testing
What is benchmarking
Treat clients in naturalistic treatment settings with a treatment program
What are practice research networks
Networks that assemble an infrastructure that involves the collaboration of clinical researchers and full time practitioners expressly to examine the generalization and ability to disseminate newly developed treatments
What are 4 reasons why replication continues
- Decentralized – not just doing same big Ex over and over, do smaller ones
- Professional reinforcement
- Close to clinic – client receives treatment
- Series are cumulative
lag
past values
