The Immediate Effects of Being Observed

1
The Immediate Effects of Being Observed

• John Austin Western Michigan University
• Don Rohn Nissan North America, Inc.

Special thanks to Dr. James Carr, Dr. Debra
Hazel, Alicia Alvero, Heather Frederick, Angela
Lebbon, Kristen Rost, Joe Sasson, Siggi
Sigurdsson This research was supported in part
by a WMU Graduate College Research and Travel
Award.
2
Durability

• Applied behavior analysts can change behavior
• Getting lasting change is a different story
• Durability, or maintenance, of behavior change is
  an important issue
• For this talk, they are interchangeable terms

3
Maintenance

• Typically conceptualized as
• Continued effects first attributed to the
  treatment
• Also
• Generalization over time (Stokes Baer, 1977)
• Continuation of effects once intervention is
  withdrawn (Boyce Geller, 2001)
• Continuation of effects (Malott Suarez, 2001)

4
(No Transcript)
5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
The Issue

• Not so much which conceptualization is correct
  as it is how to produce lasting change.
• Recent review of the OBM literature, Sigurdsson
  (2004) found high positive correlations between
  certain institutionalization activities and
• Effect size (d) between baseline and
  intervention, and
• Effect size (d) between baseline and maintenance
  or follow-up data

9
Institutionalization Activities

• Internal staff involvement in intervention design
• Internal staff trained to implement any part of
  the intervention
• Internal staff involved in formal data collection
  systems
• Internal staff involved in formal consequence
  delivery systems
• From Sigurdsson, 2004 based on McSween
  Matthews (2001) and Grindle, Dickinson,
  Boettcher (2000).

10
Aims

• Examine reactivity to observation
• What do people do when no one is looking?
• In implementing behavioral systems, this is an
  important issue
• Does feedback maintain behavior change in the
  absence of the observer?
• Secondary Aim
• What implications does this have for ABA research
  involving direct and ostensible observation?

11
Empirical Demonstrations of Reactivity

• Work Behavior
• Belfiore, Mace, and Browder (1989)
• Workers produced more work units when an observer
  was present
• Rusch et al. (1984)
• Dishwashers spent more time on task when observer
  was present
• Safety Behavior
• Olson Austin (2001)
• Bus operators responded systematically to the
  presence of a supervisor

12
Purpose

• To determine the momentary effects of a safety
  observation by examining the safe behavior of
  participants prior to, during, and immediately
  following observation.
• In the context of this symposium
• To examine the durability of behavior change on a
  rather molecular scale

13
Method

• Participants
• 6 undergraduate psychology students
• Setting
• Simulated office environment
• Observation room outfitted with a desk, desktop
  computer, adjustable chair, adjustable keyboard,
  and document holder
• Equipment adjusted prior to each session to
  remove factors that would prevent safe
  performance
• Participants engaged in a typing task (i.e.,
  transcription)

14
Method

• Dependent variables
• Neck position
• Shoulder position
• Back posture
• Wrist position
• Leg position
• Feet flat on floor
• Percentage of safe intervals calculated for each
  DV

15
Method

• Measurement of Dependent Variables
• Each 25-minute work session videotaped via hidden
  camera
• Videotapes were scored using 20-second momentary
  sampling
• Each DV scored as safe or unsafe
• Data plotted in 5-minute blocks for each session
• Mean of all pre-observation, observation, and
  post-observation blocks, across phases
• Average safe performance for each 5-minute block

16
Method

• Multi-element within subjects design with
  multiple baseline across behaviors
• Information Phase
• Participants reviewed safety handout with
  definitions of DVs prior to each session
• Participants demonstrated each behavior to
  researcher to ensure they knew how to correctly
  perform
• Overt Observation
• Observer present for 5-minute observation (no
  observer during remaining 20 min)
• Announced DVs measured
• No Feedback
• Feedback
• Observer delivered feedback to participant
  following 5-min observation period

17
General Results

• All participants demonstrated reactivity to the
  presence of the observer on at least 1 dependent
  variable
• Feedback did not appear to maintain performance
  in post-observation intervals
• In most cases, performance deteriorated
  immediately following observation
• Observation generally resulted in increases in
  safety performance for several DVs
• Data

18
Organization of Results
19
Participant A
20
Participant B
21
Participant C
22
Discussion

• Do workers perform more safely when an observer
  is present compared to working alone?
• These data suggest they do, when they are
  informed what is being observed
• Does performance deteriorate in post observation
  intervals?
• The data suggest that it doeswithin minutes

23
Discussion

• Is performance improvement a product of
  reactivity or feedback?
• The data suggest reactivity to be responsible for
  improvement
• In cases where feedback improved performance, in
  was only observed when the observer was present
• Effectiveness of feedback may have been limited
  by brevity of the intervention
• Reactivity adds additional support to
  employee-driven initiatives
• Train all employees to conduct observations

24
Discussion

• Reinforcement (or other Tx) in addition to
  feedback may improve performance in
  post-observation intervals
• To promote durability
• Taking advantage of reactivity
• Use the observation process to create reactivity
• Reactivity gives observers an opportunity to
  provide reinforcement
• The reactivity effects observed in this study
  were not short-term effects (20 sessions in some
  cases)

25
Future Research

• We should be conducting well-constructed
  experiments to better understand this phenomenon.
• Some of these are
• What role is served by contingencies that are
  indiscriminable across phases (i.e.,
  generalization)? (Boyce Geller, 2001)
• What role does involvement serve? (Boyce
  Geller, 2001)
• What role does clear and consistent behavioral
  programming play? (Malott, 2001)
• What role do rules play? (McSween Matthews,
  2001)

26

• Questions?
• john.austin_at_wmich.edu
• Join the OBM Network online at www.obmnetwork.com
  !

Special thanks to Dr. James Carr, Dr. Debra
Hazel, Alicia Alvero, Heather Frederick, Angela
Lebbon, Kristen Rost, Joe Sasson, Siggi
Sigurdsson This research was supported in part
by a WMU Graduate College Research and Travel
Award.
27

• Extras

28
Special Interventions

• Two participants were exposed to special
  interventions
• Based on poor performance on one or more DVs
• Cant do and wont do
• Participant D was exposed to corrective feedback
• Participant E was allowed to escape the work
  session early if he met safety goal

29
Limitations of Overt Observation

• Safety consultants recommend having observers
  announce when observations will occur (Krause,
  1997 McSween, 1995)
• Disadvantage inherent to this approach is that
  the observer may not see the typical work
  practices of the employee
• Reactive effect of observation
• Workers may perform systematically to the
  presence or absence of the observer
• Is this a bad thing?

30
Reactivity and Overt Observation

• Safety data gathered from observations will not
  be an accurate representation of normal work
  practices
• Could mask potential safety hazards that are
  discovered only after an injury occurs
• The is what BBS is intended to prevent!

31
Implications

• Alvero Austin (2000) suggested practical value
  of training all employees to conduct observations
• Additional support?
• If workers perform safely only when someone is
  watching, training all employees to conduct
  observations may maintain safe behavior because
  someone is always watching

32
Implications

• Common criticism of BBS observations
• People will only work safely when someone is
  watching.
• Prior research suggests workers respond
  systematically to obvious observation
• Do workers participating in a BBS process behave
  safely only when an observer is present?
• How long does safe performance maintain following
  observation?
• Is improvement in safe performance a result of
  receiving feedback or a product of reactivity to
  the observer?