You are on page 1of 16

Journal of Applied Behavior Analysis (Fall) 2020, 53, 2287–2302 NUMBER 4

The influence of instructive feedback presentation schedule:


A replication with children with autism spectrum disorder
Casey L. Nottingham, Jason C. Vladescu, Ruth M. DeBar and
Meghan Deshais
Caldwell University

Jaime DeQuinzio
Alpine Learning Group

Instructive feedback (IF) is a modification to discrete trial instruction that may increase instruc-
tional efficiency for individuals with autism spectrum disorder. Several variations of IF have
recently been evaluated in the literature; however, few studies have assessed the effectiveness and
efficiency of presenting secondary targets on continuous versus intermittent presentation sched-
ules. The current study evaluated the effectiveness and efficiency of various presentation sched-
ules of secondary targets during discrete trial instruction. Specifically, we replicated and
extended Griffen et al. (1998) by comparing a condition in which secondary targets were pres-
ented during each trial of a session, a condition in which secondary targets were presented every
other trial, and a condition in which secondary targets were presented about every 4 trials.
Within-subject replications were included for both participants. One of the intermittent presen-
tation schedules was associated with the most optimal outcomes in all 4 comparisons.
Key words: autism spectrum disorder, instructional efficiency, instructive feedback, secondary
targets, tacts

Instructive feedback (IF) is a variation of dis- The effectiveness and efficiency of IF has
crete trial teaching that involves presenting been evaluated. Effective procedures are proce-
additional, nontarget stimuli within learning dures that result in desired learning outcomes
trials (Reichow & Wolery, 2011). When using while efficient procedures are procedures that
IF, practitioners directly teach primary targets result in optimal learning outcomes without
and also present additional stimuli (hereafter increasing instructional time (Reichow &
referred to as secondary targets) that are not Wolery, 2011). The IF procedure has been
directly taught (Vladescu & Kodak, 2013). effectively used to promote skill acquisition in
Learners are not required to respond to second- 1:1 (e.g., Carroll & Kodak, 2015), dyad
ary targets and, if a response is given, the prac- (e.g., Appelman et al., 2014), and small group
titioner withholds feedback related to the (e.g., Cromer et al., 1998; Leaf et al., 2017)
response. formats, and with individuals with a range of
disabilities, including autism spectrum disorder
This project was completed in partial fulfillment of the
(e.g., Appelman et al., 2014; Caldwell et al.,
requirements for a Doctor of Philosophy in Applied 1996; Carroll & Kodak, 2015; Cromer et al.,
Behavior Analysis by the first author, under the supervi- 1998; Dass et al., 2018; Delmolino et al.,
sion of the second author. 2013; Haq et al., 2015; Parrott et al., 2000;
Address correspondence to: Jason C. Vladescu, Depart-
ment of Applied Behavior Analysis, Caldwell University, Ross & Stevens, 2003; Schnell et al., 2018;
120 Bloomfield Avenue, Caldwell, NJ 07006, Tekin-Iftar et al., 2003; Tullis, Marya, &
Email: jvladescu@caldwell.edu or Casey L. Nottingham, ABA Shillingsburg, 2019; Werts et al., 2011).
Collective, LLC., 1 International Boulevard Ste 400, Mahwah,
NJ 07495. Email: cnottingham@abacollective.com The IF procedure has also proven effective
doi: 10.1002/jaba.706 across several procedural variations such as the
© 2020 Society for the Experimental Analysis of Behavior
2287
2288 Casey L. Nottingham et al.

location of secondary targets within a learning experimenter presented secondary targets either
trial. Secondary targets have been presented during every trial of a session (i.e., continuous
within the antecedent portion (e.g., Haq et al., presentation) or about every four trials
2015; Vladescu & Kodak, 2013) of the trial (i.e., intermittent presentation). Participants
prior to the presentation of a primary target. acquired the primary and secondary targets in
Secondary targets have also been presented both conditions. The researchers reported that
within the consequence portion the intermittent presentation condition
(e.g., Carroll & Kodak, 2015) of the trial fol- required less training time than the continuous
lowing reinforcement for a correct response presentation condition.
related to a primary target. Lastly, secondary Within this study, Griffen et al. (1998) eval-
targets have been embedded within the ante- uated IF procedures with individuals with mod-
cedent stimulus related to the primary target erate intellectual disabilities to teach tacts as the
(e.g., Doyle et al., 1996) of a learning trial (see primary targets (e.g., Lexington Public Library)
Nottingham et al., 2015 for examples). and additional information about the primary
Previous research has also evaluated the targets as secondary targets (e.g., “Lexington
effects of presenting multiple secondary tar- Public Library is where we check out books”).
gets within each learning trial (e.g., Gast Secondary target probes were conducted as a
et al., 1994; Nottingham et al., 2017). For two-phase procedure. During Phase 1 of the
example, two secondary targets have been secondary target probes, the experimenter pres-
presented within the consequence portion of ented a picture of the primary target
each learning trial (e.g., Gast et al., 1994; (e.g., Lexington Pubic Library) and said, “Tell
Nottingham et al., 2017) or one secondary me what we can do here.” If a correct response
target has been presented in the antecedent was not emitted within 5 s, the experimenter
portion and one secondary target has been progressed to Phase 2 of the secondary target
presented in the consequence portion of each probe and presented a fill-in-the-blank
learning trial (Nottingham et al., 2017). intraverbal antecedent stimulus such as, “Lex-
When compared to traditional discrete trial ington Public Library is where we can ____.”
arrangements, IF procedures involving ante- The researchers reported that all participants
cedent and consequence arrangements have acquired the secondary target responses in the
been demonstrated to result in better instruc- absence of direct teaching; however, data for
tional efficiency by increasing the number of secondary target probes were combined across
targets acquired without substantially increas- Phases 1 and 2. The combined summarization
ing instructional time (e.g., Nottingham of data prevents conclusions regarding
et al., 2017; Vladescu & Kodak, 2013). (a) whether one secondary target probe ante-
Another variation of the arrangement of IF cedent stimulus resulted in higher levels of cor-
that may affect both the effectiveness and effi- rect responding and (b) if the findings were
ciency of instruction involves the schedule or consistent across participants. Additional
frequency of presentation of secondary targets. research is needed to assess the generality of the
Secondary targets may be presented on a con- findings of Griffen et al. to different skills and
tinuous (i.e., during each trial of a session) or different learners. For example, individuals with
intermittent (i.e., during some trials of a ses- ASD may benefit from discrete trial instruction
sion) basis. Griffen et al. (1998) evaluated the that incorporates these procedural modifica-
continuous and intermittent presentation of tions. Another potential expansion related to
secondary targets with participants diagnosed this research area (Griffen et al., 1998) involves
with moderate intellectual disabilities. The including a presentation schedule that allows
Secondary Target Presentation Schedule 2289

for intermittent presentation of secondary tar- principles of applied behavior analysis for at
gets but does not restrict presentation of those least 5 years at the start of the study. At the
targets to once per session (e.g., presentation of beginning of the study, both participants dem-
secondary targets twice per session). Individuals onstrated appropriate learning behavior
diagnosed with ASD who may require multiple (e.g., sitting at a desk, orienting towards the
exposures to a target for the establishment of instructor, attending to instructional stimuli);
stimulus control may benefit from a modified were considered to have generalized imitative
intermittent presentation schedule of secondary and echoic repertoires; emitted at least 1,000
targets; however, no research has been con- tacts; vocally manded for items, information,
ducted evaluating differing intermittent presen- and the termination of aversive events; and
tation schedules of secondary targets. demonstrated intraverbals under a variety of
The purposes of the current study were to stimulus control categories (simple, compound,
systematically replicate and extend Griffen et al. verbal conditional discriminations; Sundberg,
(1998). The study replicated Griffen et al. by 2016). Both participants received low scores on
evaluating (a) a continuous presentation sched- the Barriers Assessment of the Verbal Behavior-
ule of secondary targets during which a second- Milestones Assessment and Placement Program
ary target was presented following every (Sundberg, 2008) (6 for John, 17 for Dennis),
primary target in a session and (b) an intermit- and demonstrated all skills on Levels 1–3 from
tent schedule of secondary targets during which the tact domain and Levels 1–2 of the echoic
a secondary target was presented every four tri- domain. Prior to the current study, no addi-
als during a session. The study extended tional norm- or criterion- referenced tests were
Griffen et al. by (a) including an additional administered to either participant. Neither par-
intermittent condition during which a second- ticipant had any substantial history with IF.
ary target was presented every other trial in a Three additional participants started the
session, (b) evaluating various presentation evaluation; however, data for two participants
schedules with children with ASD, and were not included because they developed dis-
(c) including different measures of instructional ruptive behavior that interfered with session
efficiency that have been more commonly used completion and the acquisition of primary and
in recent IF research. secondary target responses. Data for the third
participant was not included because a replica-
tion comparison could not be conducted.
Method
Participants
Two children diagnosed with ASD by an Settings and Materials
independent professional participated. Parents The experimenter conducted all sessions in
of all participants completed the Gilliam an area of each participant’s home designated
Autism Rating Scale-Third Edition (Gilliam, for the study. Denis’ sessions took place at a
2013) to document characteristics associated table with two chairs in a bedroom. John’s ses-
with an ASD diagnosis. Scores indicated that a sions took place in a seating area in the living
diagnosis of ASD was very likely. John was room. Both participants had prior histories of
7 years, 6 months old and had received inter- instruction in these locations. During all ses-
vention based on the principles of applied sions, data sheets, picture stimuli, timers, indi-
behavior analysis for at least 4 years at the start vidualized token systems, and a video camera
of the study. Denis was 8 years, 6 months old were present. Picture stimuli cards were either
and had received intervention based on the 7.62 by 12.7 cm (Denis) or 10.16 by
2290 Casey L. Nottingham et al.

15.24 cm (John) and depicted targets that were attending during the first trial of a session, and
individualized for each participant. The experi- was stopped following completion of the final
menters recorded all sessions using a video discrete trial component (e.g., reinforcement,
camera. presentation of a secondary target) of the last
trial for the session. Reinforcer consumption
duration was excluded from total session dura-
Experimental Design and Response tion by pausing the session duration timer
Measurement when the participant earned the final token of
An adapted alternating treatments design the token system. The session duration timer
(Sindelar et al., 1985) with a no-treatment con- remained paused throughout the reinforcement
trol condition was used to evaluate the effec- interval. The session duration timer was then
tiveness and efficiency of three secondary target restarted immediately prior to establishing
presentation schedules to teach tacts to individ- attending for the next trial.
uals with ASD. During all sessions, the experi- The percentage of unprompted correct
menter recorded data using data sheets created responses is depicted in the figures and was cal-
for each condition. Following the identification culated for each session and across conditions
of primary and secondary targets for each par- by dividing the number of unprompted correct
ticipant, data sheets were individualized with responses by the total number of trials in the
primary and secondary targets and target order session and multiplying the resulting ratio by
was randomized using a random without 100. To evaluate instructional efficiency, we
replacement method. During sessions, data calculated the mean training time per target
were collected on unprompted correct across conditions by dividing the total training
responses, prompted correct responses, time for each condition by the number of tar-
unprompted incorrect responses, prompted gets (i.e., primary and secondary) mastered
incorrect responses, and echoics. Unprompted within the condition.
correct responses were defined as a participant
emitting a predefined response corresponding
to the antecedent stimulus prior to the experi- Interobserver Agreement
menter’s delivery of a prompt. Prompted cor- A second independent observer collected
rect responses were defined as a participant data during 33% of baseline sessions across
emitting a predefined response corresponding conditions for John and Denis’ initial and repli-
to the antecedent stimulus following the experi- cation evaluations for interobserver agreement
menter’s delivery of a prompt. Unprompted (IOA) purposes. Similarly, data were collected
incorrect responses were defined as a partici- during 50% of treatment comparison sessions
pant emitting an error of commission or omis- across conditions for John’s initial and replica-
sion prior to the experimenter’s delivery of a tion evaluations, and 33% of treatment com-
prompt. Prompted incorrect responses were parison sessions across conditions for Denis’
defined as a participant emitting an error of initial and replication evaluations. For trial-by-
commission or omission following the experi- trial IOA data, an agreement was defined as
menter’s delivery of a prompt. Echoics were any instance in which the primary and second-
defined as imitation of the experimenter’s ary observers scored a trial in the same way. A
model of the secondary target within 3 s. In disagreement was defined as any instance in
addition, the experimenter recorded session which the primary and secondary observers
duration using a digital timer. The timer was scored a trial differently. IOA was calculated by
started immediately prior to establishing dividing the number of agreements by the
Secondary Target Presentation Schedule 2291

number of agreements plus disagreements and integrity was calculated by dividing the total
multiplying the resulting ratio by 100. For ses- number of correct experimenter behaviors by
sion duration IOA data, the shorter duration the number of correct experimenter behaviors
recorded by the two observers was divided by plus the number of incorrect experimenter
the longer duration and the resulting ratio was behaviors and multiplying the resulting ratio by
multiplied by 100. 100. For both John’s and Denis’ initial and
For John’s initial evaluation, IOA was 100% replication evaluations, mean procedural integ-
for all conditions as well as for echoic rity was 100% during baseline and treatment
responding across conditions. Mean duration comparison phases across all conditions.
IOA was 92% (range, 78%–100%) across the A second independent observer recorded
continuous, intermittent 4:1, and intermittent procedural integrity during 100% of the proce-
4:2 conditions. For John’s replication evalua- dural integrity sessions to evaluate IOA related
tion, IOA was 100% across all conditions. to procedural integrity measurement. The IOA
Echoic IOA was 100% for the intermittent 4:1 score was calculated by dividing the number of
and intermittent 4:2 conditions. Mean echoic agreements by the number of agreements plus
IOA was 95% (range, 90%–100%) for the disagreements and multiplying the resulting
continuous condition. For John’s replication ratio by 100. For both John and Denis, IOA
evaluation, mean duration IOA was 92% on procedural integrity was 100% across all
(range, 79%–100%). conditions in the baseline and treatment com-
For Denis’ initial evaluation, IOA was 100% parison phases for both the initial and replica-
for all conditions as well as for echoic tion evaluations.
responding across conditions. Mean duration
IOA was 93% (range, 90%–95%) across condi-
Preference Assessments and Token
tions. For Denis’ replication evaluation, IOA Economy
was 100% across all conditions as well as for
We conducted paired-stimulus preference
echoic responding across conditions. For Denis’
assessments (Fisher et al., 1992) prior to the
replication evaluation, mean duration IOA was
onset of the study for each participant. Either
94% (range, 92%–97%).
an edible or tangible preference assessment was
conducted. Any item selected at least one time
during the preference assessment was available
Procedural Integrity and Procedural
during sessions. These items were available for
Integrity Interobserver Agreement
selection following exchange of 10 tokens, pro-
A trained observer collected data on proce-
vided contingent on prompted correct and
dural integrity during at least 33% of sessions
unprompted correct responses, or appropriate
across conditions for each participant using
sitting and attending responses throughout the
data sheets created for each condition. The
study. Any item vocally manded for by a partic-
observer recorded whether the procedures were
ipant that was not included in the preference
implemented as planned for each condition by
assessment was also available.
recording whether the experimenter correctly or
incorrectly established attending behavior, pres-
ented the primary and secondary targets as Target Identification and Assignment
applicable, delivered prompts, provided or For each participant, the experimenter con-
withheld feedback for responses to primary and ducted pretests to identify potential primary
secondary targets, and waited for the intertrial and secondary targets. Potential targets were
interval prior to starting a new trial. Procedural identified based on individualized treatment
2292 Casey L. Nottingham et al.

goals, caregiver suggestions, or participant correct responses, the experimenter increased


mands, and included tact targets for the initial the prompt delay to 5 s. Similar to previous IF
and replication phases. studies (Nottingham et al., 2017; Vladescu &
During pretest trials, the experimenter pres- Kodak, 2013), the experimenter implemented a
ented the antecedent stimulus and allowed par- re-present-until-independent error correction
ticipants 5 s to respond. The experimenter procedure following unprompted (5-s prompt
withheld feedback for unprompted correct and delay) and prompted (0-s prompt delay) incor-
unprompted incorrect responses but provided a rect responses. The experimenter provided a
token for appropriate sitting and attending vocal prompt of the correct response and
(e.g., hands in lap, feet on floor, sitting still, allowed an opportunity for the participant to
establishing eye contact with the experimenter) engage in the correct response. If the partici-
about every three trials, during the intertrial pant engaged in a prompted correct response,
interval. Each potential target was pretested the experimenter provided a statement of
three times. If a participant responded correctly acknowledgement (e.g., “Right”) and re-
to a potential target at least one time, it was presented the same trial until the participant
discarded. The remaining targets were entered engaged in an unprompted (5-s prompt delay)
into a pool of targets that could potentially be or prompted (0-s prompt delay) correct
assigned to various conditions across phases of response. Error correction data are not included
the study using a logical analysis (Wolery et al., in the figures. During each treatment compari-
2014). Targets that were physically or phoneti- son condition, prompted (0-s prompt delay) or
cally similar were not included within the same unprompted (5-s prompt delay) correct
set, and the number of syllables per target were responses to the primary target resulted in
equated across conditions. Targets that began praise and the delivery of a token. During trials
with the same letter were not included within in which secondary targets were presented, the
the same set. For each comparison, we assigned experimenter held up the target picture for 3 s
six targets to the continuous condition (three and said, “This is a (stated name).” Participants
primary targets plus three secondary targets), were not required to respond to these targets
six targets to the intermittent conditions (three and if a response (e.g., echoic) was given, no
primary targets plus three secondary targets), feedback was provided by the experimenter.
and three targets to each control conditions. The mastery criterion for primary and second-
See Table 1 for a list of targets. ary targets was two consecutive sessions with
unprompted correct responding at 100%. Each
comparison was conducted twice with different
General Procedures stimuli for replication purposes.
The experimenter conducted sessions 1 to
4 days per week based on participant and Baseline and Control
experimenter availability. During sessions in During baseline and control sessions, the
conditions involving teaching of primary tar- experimenter established attending, presented
gets, vocal model prompts were used and faded the antecedent stimuli (held up a picture and
using a constant prompt delay procedure. Dur- said, “What is it?”), and allowed the participant
ing the first two instructional sessions, the 5 s to respond. The experimenter did not pro-
experimenter presented vocal prompts immedi- vide feedback for unprompted correct or
ately following the delivery of the antecedent unprompted incorrect responses during baseline
stimuli (i.e., 0-s prompt delay). Following two sessions. Praise and a token were delivered
consecutive sessions with 100% prompted every other trial for appropriate attending and
Secondary Target Presentation Schedule 2293

Table 1

Primary and Secondary Targets

Participant Primary Control Secondary Control CONT INT 4:2 INT 4:1
John (initial) Ana Aengus Carthage Fitz Roy Brigid
Red Beach Tamarind Starry Night Jane Goodall Melting Watch
Silver Surfer Asparagus Lotus Temple The Yellow House Praying Mantis
Morrigan Krishna Dagda
The Thinker Irises Taj Mahal
Paper Towels Golden Temple 100 Cans
John* Ohio Delaware Angel Falls Singapore Angkor Wat
Barcelona Copenhagen Pennsylvania Table Mountain North Dakota
Mexico City Philadelphia Tennessee State Flag Oregon Hawaii State Flag
Istanbul Rhode Island Vienna
Massachusetts Abu Dhabi Connecticut
New Hampshire Michigan Illinois
Denis (initial) Manatee Sea Urchin Guinea Pig Garbage Truck Jellyfish
Harvester Cable Car Katydid Centipede Mobile Home
Stink Bug Earwig Swordfish Ferret Pitchfork
Pelican Tambourine Wheelbarrow
Clarinet Screwdriver Scorpion
Forklift Puffin Golf Cart
Denis* Squid Plough Tie Screw Swan
Flute Lobster Nails Eel Worm
Artichoke Calculator Grasshopper Mosquito Hermit Crab
Scarf Clam Vase
Cockroach Beetle Urchin
Fire Hydrant Paper Towels Praying Mantis

Note. Italicized words indicate secondary targets. Asterisks (*) indicate replication treatment comparison.

sitting behaviors to approximate the schedule presented following any primary target. This
of reinforcement in effect during intervention condition was similar to the continuous presen-
conditions. tation condition in Griffen et al. (1998).

Continuous Presentation Intermittent 4:2 Presentation


During each trial in the continuous During each trial in the intermittent (INT)
(CONT) presentation condition, a secondary 4:2 presentation condition, each primary target
target was presented every time a primary target was presented four times and each secondary
was presented. During each trial, the experi- target was presented two times. During each
menter established attending and presented the trial within this condition, the experimenter
antecedent stimuli relevant to the primary tar- established attending and presented the ante-
get. During this condition each of the three cedent stimulus relevant to the primary target.
primary targets was presented four times for a Each of the three primary targets was presented
total of 12 trials, and each of the three second- four times for a total of 12 trials, and each of
ary targets was presented four times for a total the three secondary targets was presented two
of 12 presentations, using a random without times for a total of six presentations using a
replacement selection method for determining random without replacement selection method
the order of targets within a session. Primary for determining the order of targets within a
and secondary targets were not paired session. Primary and secondary targets were not
(i.e., linked) when presented during this condi- paired (i.e., linked) when presented during this
tion; that is, any secondary target could be condition.
2294 Casey L. Nottingham et al.

Intermittent 4:1 Presentation condition was similar to the intermittent sched-


During each trial in the INT 4:1 presentation ule condition from Griffen et al. (1998).
condition, each primary target was presented four
times and each secondary target was presented Secondary Target Probes
one time per session. During each trial, the Probes were conducted immediately prior to
experimenter established attending and presented every instructional session to evaluate partici-
the antecedent stimulus relevant to the primary pants’ acquisition of secondary targets. Secondary
target. Procedures during this condition were target probe procedures were identical to the pro-
identical to the INT 4:2 presentation condition; cedures outlined in the baseline and control con-
however, during the INT 4:1 presentation condi- ditions. The experimenter established attending,
tion, each of the three primary targets was pres- presented the antecedent stimulus relevant to the
ented four times for a total of 12 trials and each secondary target, allowed 5 s for the participant
of the three secondary targets was presented one to respond, withheld feedback for unprompted
time for a total of three presentations using a correct and incorrect responses, and delivered
random without replacement selection method praise and a token about every trial for appropri-
for determining the order of targets within a ses- ate sitting and attending behavior during the
sion. Primary and secondary targets were not intertrial interval. Although secondary targets
paired (i.e., linked) during this condition. This were presented at varying presentation schedules

Figure 1
John’s Unprompted Correct Responses across Conditions
Primary Targets Primary Targets
BL Treatment Comparison BL Treatment Comparison
100 100
Primary Control
80 80 Continuous
Percentage of Unprompted Correct Responses

INT 4:1
60 60 INT 4:2
Echoics
40 40

20 20

0 0
0 20 40 60 0 20 40 60
Sessions Sessions

Secondary Targets
Secondary Targets
BL Secondary Target Probes
BL Secondary Target Probes
100 100
Secondary Control
80 80 Continuous
INT 4:1
60 60 INT 4:2

40 40

20 20
John (initial) John (replication)
0 0
0 20 40 60 0 20 40 60
Sessions Sessions
Secondary Target Presentation Schedule 2295

during instructional sessions (i.e., CONT, INT Results


4:2, INT 4:1), during secondary target probe ses- Figures 1 and 2 depict results for John and
sions, each secondary target was presented four Denis. In each graph, the percentage of
times across each condition. unprompted responses related to primary and
secondary targets are depicted in the top and
Review Sessions bottom panels, respectively. During baseline,
Review sessions were conducted for pri- the percentage of unprompted correct responses
mary targets while secondary target probes was at or near zero for both primary and sec-
continued if primary targets were mastered, ondary targets for both participants. Addition-
but secondary targets were approaching mas- ally, unprompted correct responding in the
tery. This arrangement allowed for an evalua- primary and secondary control conditions
tion of whether additional exposure would remained at or near zero for both participants
result in mastery of the secondary targets. across phases.
Review sessions would have also been con- Figure 1 displays results for John’s initial and
ducted, although were not necessary, for sec- replication comparisons. During the treatment
ondary targets as needed while continued comparison phase, John acquired the primary
teaching to mastery was ongoing for primary targets in all three instructional conditions in
targets in a condition. four training sessions. He engaged in correct

Figure 2
Denis’ Unprompted Correct Responses across Conditions
Primary Targets Primary Targets
BL Treatment Comparison BL Treatment Comparison
100 100

80 Primary Control
Percentage of Unprompted Correct Responses

80
Continuous
60 60 INT 4:1
INT 4:2
40 40 Echoic

20 20

0 0
20 40 60 80 100 20 40 60
Sessions Sessions

Secondary Targets Secondary Targets


BL Secondary Target Probes BL Secondary Target Probes
100 100

80 80 Secondary Control
Continuous
60 60 INT 4:1
INT 4:2
40 40

20 20
Denis (initial) Denis - Replication
0 0
20 40 60 80 100 20 40 60
Sessions Sessions
2296 Casey L. Nottingham et al.

Figure 3
Mean Training Time
150 150
2 min
1 min 1s
100

S eco n d s
100
S eco n ds

44 s
1 min 19 s 1 min 13 s 1 min 14 s 1 min 8 s

50 50

John (initial) John (replication)


0 0

us

1
4:
4:
1

2
us

uo
4:

4:

T
T
uo

tin

IN
IN
T

T
tin

on
IN

IN

C
on

C ondition
C

C ondition

200
200
2 min 36 s
150
2 min 1 min 59 s 150 2 min
Seconds

20 s
Seconds

100 1 min 48 s
100
1 min 15 s
50
50
Denis (initial) Denis (replication)
0 0
us

2
4:

4:

us

1
uo

4:

4:
uo
T

T
tin

T
IN

IN

in

IN

IN
on

nt
Co
C

C on d ition Condition

echoic responding following the presentation of CONT, INT 4:2, and INT 4:1 conditions.
secondary targets during a mean of 98% of tri- During secondary target probes, John again
als (range, 92%-100%) in the CONT condi- demonstrated mastery of the secondary targets
tion and 100% of trials in the INT 4:2 and in all conditions in three sessions, prior to mas-
INT 4:1 conditions. John demonstrated mas- tery of the primary targets in all conditions.
tery of the secondary targets in all conditions in Figure 2 displays results for Denis’ initial
three secondary target probe sessions, prior to and replication comparisons. During the initial
demonstrating mastery of the primary targets in treatment comparison, Denis demonstrated
all conditions. For John’s replication compari- mastery of the primary targets in the CONT
son, during the treatment comparison phase, and INT 4:2 conditions in five sessions each
he demonstrated mastery of the primary targets and in the INT 4:1 condition in six sessions.
in the INT 4:1 and INT 4:2 conditions in four Denis engaged in correct echoic responding
sessions each. He demonstrated mastery of the during 100% of trials across the CONT, INT
primary targets in the CONT condition in five 4:2, and INT 4:1 conditions. During secondary
sessions. During instructional sessions, John target probes of the initial treatment compari-
engaged in correct echoic responses following son, Denis did not demonstrate mastery-level
the experimenter’s presentation of the second- responding to the secondary targets prior to
ary target during 100% of trials across the mastering the primary targets; however,
Secondary Target Presentation Schedule 2297

increasing or variable trends were observed in conditions, respectively. The mean training
all conditions at the time primary targets were times per target for Denis’ replication compari-
mastered in the corresponding conditions. The son were 2 min 20 s, 1 min 48 s, and 1 min
experimenter conducted review sessions for the 15 s for the CONT, INT 4:2, and INT 4:1
primary targets while continuing secondary tar- conditions, respectively.
get probes to determine whether additional
exposure would result in mastery of the second-
ary targets. Denis demonstrated mastery of the Discussion
secondary targets in the INT 4:1 condition in The current study evaluated the effectiveness
nine probe sessions, in the CONT condition in and efficiency of presenting secondary targets
ten probe sessions, and in the INT 4:2 condi- on continuous and intermittent presentation
tion in 12 probe sessions. schedules with two participants diagnosed with
For Denis’ replication comparison, during ASD. Both participants acquired the primary
the treatment comparison, Denis mastered the targets using a constant prompt delay proce-
primary targets in the INT 4:1 condition in dure, reinforcement, and a re-presentation-
four sessions and in the INT 4:2 and CONT until-independent error correction teaching
conditions in five sessions each. Denis engaged procedure, with the incorporation of secondary
in correct echoic responding during 100% of targets. Both participants also acquired the sec-
trials across the CONT, INT 4:2, and INT 4:1 ondary targets in the absence of direct teaching,
conditions. During secondary target probes, suggesting that IF was an effective procedure in
Denis acquired the secondary targets in the this study. Similar to the findings of Griffen
INT 4:2 condition in four probe sessions, prior et al. (1998), the current study found that
to mastery of the primary targets in that condi- intermittent presentation of secondary targets
tion. He acquired the secondary targets in the during IF was more efficient than a continuous
CONT condition in five probe sessions, equal presentation schedule of secondary targets. The
to the number of instructional sessions that current study allowed for further investigation
were required for him to acquire the primary of intermittent presentation schedules by
targets in that condition. He acquired the sec- including both the INT 4:1 and INT 4:2
ondary targets in the INT 4:1 condition in schedules. Across all four comparisons with
seven probe sessions, after mastery of the pri- both participants, results showed that one of
mary targets. the intermittent presentation conditions was
Figure 3 depicts mean training time per tar- associated with the lowest mean training time
get across instructional conditions for John’s per target. However, differences (albeit, mini-
and Denis’ initial and replication evaluations. mal) were observed regarding which intermit-
The mean training times per target for John’s tent presentation schedule resulted in the most
initial evaluation were 1 min 44 s, 1 min 19 s, efficient acquisition. Importantly, experi-
and 1 min 13 s for the CONT, INT 4:1, and menters also found that there was no detrimen-
INT 4:2 conditions, respectively. The mean tal impact on participants’ acquisition of
training times per target for John’s replication primary targets regardless of whether secondary
evaluation (Figure 3) were 2 min 1 s, 1 min targets were presented continuously or
14 s, and 1 min 8 s for the CONT, INT 4:2, intermittently.
and INT 4:1 conditions, respectively. The Although measures were taken to equate
mean training times per target for Denis’ initial stimulus sets across conditions in each treat-
evaluation were 2 min 36 s, 2 min, and 1 min ment comparison, differences among target sets
59 s for the CONT, INT 4:1, and INT 4:2 may have contributed to which intermittent
2298 Casey L. Nottingham et al.

condition resulted in the least amount of train- Nottingham et al., 2017; Dwight and Rick
ing time per target. Regardless, an intermittent from Vladescu & Kodak, 2013), John acquired
presentation of secondary targets (i.e., INT 4:2 the secondary target responses prior to demon-
or INT 4:1) resulted in the least amount of strating mastery of the primary target responses
training time per target and the most efficient in both the initial and replication evaluations.
acquisition across all participants and treatment This outcome could be a function of the differ-
comparisons. Researchers should attempt to ence in the minimum training sessions required
replicate this finding and determine the condi- before participants could demonstrate mastery-
tions under which intermittent presentation level responding (two consecutive sessions with
schedules produce optimal outcomes. Partici- 100% unprompted correct responding) for pri-
pant characteristics should also be further mary and secondary targets. For primary tar-
explored to better understand the learners for gets, the first two training sessions were
whom intermittent schedules will be effective conducted with a 0-s prompt delay. Thereafter,
when using IF. The current results should the prompt delay was increased to 5 s. Thus, a
be considered in light of the participants’ well- minimum of four training sessions were
developed echoic, tact, and intraverbal reper- required before it was possible for participants
toires. Practitioners should consider the similar- to demonstrate mastery of primary target
ity of clients when applying this type of IF responses. However, because the experimenter
schedule in clinical practice given the lack of initiated secondary target probes prior to the
clarity regarding minimally required repertoires first instructional session, a minimum of three
and limited research within this area. Future secondary target probe sessions would seem-
replications should include a primary target ingly be required before it would be possible
only condition to evaluate the extent to which for participants to demonstrate mastery of sec-
intermittent presentations of secondary targets ondary target responses. Researchers could
further increases instructional efficiency over attempt to determine the conditions under
standard discrete trial instruction. which participants acquire secondary target
It should be noted that the preparation did responses quicker than primary target
not allow for an evaluation of the influence of responses. For example, it might be useful to
presentation schedule distinct from the mini- eliminate the 0-s prompt delay training ses-
mum number of exposures necessary to achieve sions. That is, following baseline, training
mastery. That is, the number of exposures could start with a 5-s prompt delay. For certain
rather than the ratio of primary to secondary participants, and with a particular subset of cli-
targets may play a role in acquisition of second- ents in clinical practice, this variation would
ary target responses. In this study, the INT make it possible for participants to demonstrate
conditions could have provided sufficient expo- mastery of primary and secondary target
sures for secondary target responses to be responses following the same minimum num-
acquired whereas the CONT condition pro- ber of training sessions. For example, future
vided more exposures than necessary. Future researchers or practitioners working with indi-
researchers could conduct evaluations to evalu- viduals who readily acquire targeted responses
ate the distinct influence of presentation sched- in the absence of errorless teaching strategies
ule while accounting for exposures. could eliminate 0-s prompt delay trials. This
The findings of the current study are similar arrange may help determine the conditions
to those of previous IF studies in several ways. under which mastery of secondary target
First, like some participants in previous IF responses prior to mastery of primary target
studies (e.g., Kelly and Simon from responses occurs. However, this remains an
Secondary Target Presentation Schedule 2299

empirical question and practitioners may led to increased instructional efficiency, the
choose to arrange clinical teaching practices by current study adds to the small body of litera-
individualizing instructional strategies best ture (e.g., Nottingham et al., 2017) that has
suited for their clients to prevent errors early in shown variations of IF hold promise as a more
learning. optimal means of establishing new skills for
Second, like previous research related to IF learners with ASD. Relatedly, Nottingham
procedures (e.g., Schnell et al., 2018; et al. (2017) demonstrated that increases in
Vladescu & Kodak, 2013), the current study instructional efficiency may be achieved by pre-
presented primary and secondary targets that senting multiple (as opposed to a single) sec-
consisted of expansion targets (Reichow & ondary targets per trial. Researchers should
Wolery, 2011). Expansion targets are primary evaluate additional IF variations and combina-
and secondary targets that are either related tions of IF variations as a means to identify the
conceptually, or from the same instructional most optimal means of incorporating IF into
area. The current study evaluated expansion- acquisition programs for learners.
type targets by presenting tact targets as both Fourth, like many participants in previous
the primary and secondary targets for both par- studies (Dass et al., 2018; Nottingham et al.,
ticipants. One limitation to this arrangement is 2017; Vladescu & Kodak, 2013), the partici-
that these procedures do not generally reflect pants of the current evaluation consistently
traditional instructional arrangements within emitted overt echoics following the experi-
less restrictive teaching settings. For example, menter’s presentation of the secondary targets.
the current study taught primary targets, such Additionally, the participants consistently dem-
as a tact of a guinea pig, and then presented a onstrated behaviors consistent with attending
picture of a clarinet with the accompanying (e.g., looking at stimuli, orienting toward stim-
statement, “This is a clarinet.” Although these uli, pointing toward stimuli) to the secondary
targets are considered expansion targets because targets. The occurrence of these behaviors may
they are both tacts, the additional information suggest a naming repertoire as a mechanism rel-
included in the secondary target is not related evant to positive IF outcomes (e.g., Horne &
to the primary target with regard to the con- Lowe, 1996; Miguel, 2016).
tent, which is what may be more typical in a Participants’ demonstration of a naming rep-
less restrictive learning environment. For exam- ertoire may explain why secondary target
ple, a teacher or caregiver may similarly ask responses are acquired in the absence of direct
about a picture of a guinea pig (i.e., primary teaching. When naming, the participant acts as
target); however, it is likely that the secondary both a listener and speaker following the pre-
target—or additional information—that follows sentation of a stimulus. For example, following
will be conceptually related to the primary tar- the presentation of a secondary target, partici-
get, such as, “A guinea pig is an animal.” pants may have responded as a listener
Although some research (e.g., Dass et al., (i.e., orienting to the picture) and speaker
2018) has been conducted related to the effects (i.e., overtly or covertly echoing the name of
of IF procedures along this area of conceptual the picture) (e.g., Greer & Ross, 2008; Miguel,
expansion, future research is needed to deter- 2016). Recent IF studies have begun to incor-
mine how to best arrange these procedures for porate discussions of the potential role of nam-
individuals with ASD and for whom these ing in the effectiveness of IF procedures
types of procedures may be effective. (e.g., Dass et al., 2018; Frampton &
Third, by demonstrating that the manipula- Shillingsburg, 2019; Tullis et al., 2016). Prior
tion of a secondary target presentation schedule to the onset of the current study, naming was
2300 Casey L. Nottingham et al.

not directly assessed for either John or Denis; based on the principles of behavior analysis,
however, during the study, both participants and highly developed tact repertoires. These
consistently engaged in listener responding variables in combination may have resulted in
related to the presentation of secondary targets. fewer instructional trials required for the cur-
Like participants in recent research rent participants to demonstrate mastery of tar-
(Frampton & Shillingsburg, 2019), John and gets presented—including secondary targets—
Denis both engaged in behaviors such as resulting in more efficient mastery of targets
looking at, pointing to, or orienting toward the that were presented intermittently. Because
secondary target picture cards when the experi- limited standardized assessment data are avail-
menter presented these stimuli. Additionally, able for the current participants, future research
both participants engaged in consistent, overt is needed to replicate these procedures with
echoic responses following the experimenter’s participants of varying skill levels. Future
presentation of the secondary targets. The researchers should include detailed participant
observation of both speaker and listener characteristic profiles, particularly related to the
responses and subsequent mastery of secondary skill targeted for instruction within the study.
targets indicates the likelihood of naming reper- Several areas for future research are worth
toires. Future research related to IF should noting. First, few IF evaluations have incorpo-
incorporate evaluation of participants’ naming rated procedures to program for or assess main-
repertories prior to the onset of the study and tenance of primary and secondary targets. The
should include measurement of echoic and current study did not assess maintenance of pri-
attending behaviors to inform discussion related mary or secondary targets. Although the cur-
to the potential influence of naming rent study indicated that intermittent
repertoires. presentation schedules of secondary targets
Additional explanations for why participants resulted in more efficient learning outcomes, it
in IF studies learn secondary target responses in is unclear whether correct responding maintains
the absence of direct teaching have been pro- when acquired under intermittent presentation
posed. Indiscriminable contingencies may play schedules. Further research is necessary to eval-
a role in the effectiveness of IF procedures uate whether maintenance of acquired targets is
(Tullis, Gibbs et al., 2019). Contingencies for impacted by the secondary target presentation
responding to primary targets versus secondary schedule used during instructional sessions.
targets were likely indiscriminable across both Second, although within-subject replications
continuous and intermittent conditions. Similar were included, both targeted tacts as primary
densities of reinforcement across the teaching and secondary targets. Thus, it is unclear
sessions and secondary target probe conditions whether intermittent presentation schedules
may have resulted in participants responding would remain the more efficient procedure
similarly to both primary and secondary targets when teaching different skills (e.g., fill-in-the-
to maximize access to reinforcement through- blank intraverbals; question answering). IF pro-
out sessions. cedures have been used to teach a wide variety
Demand characteristics may also be relevant. of skills, primarily related to verbal and play-
That is, across instructional sessions and sec- based behavior. However, future researchers
ondary target probe sessions in all conditions, could include within-subject replications across
sessions were conducted in the same environ- different skill types (e.g., tacts, intraverbals,
ment by the same experimenter. Relatedly, play skills) to determine whether the most effi-
both participants demonstrated extensive histo- cient secondary target presentation schedule
ries with discrete trial instruction, instruction persists across skill types.
Secondary Target Presentation Schedule 2301

Finally, to our knowledge, this is the first 381–399. https://doi.org/10.1177/00224669960


study to evaluate secondary target presentation 2900402
Fisher, W., Piazza, C. C., Bowman, L. G.,
schedules with individuals diagnosed with Hagopian, L. P., Owens, J. C., & Slevin, I. (1992).
ASD. Currently, procedures based on the prin- A comparison of two approaches for identifying rein-
ciples of behavior analysis are widely used and forcers for persons with severe and profound disabil-
ities. Journal of Applied Behavior Analysis, 25(2),
effective for increasing skill repertoires of indi- 491–498. https://doi.org/10.1901/jaba.1992.25-491
viduals diagnosed with ASD. To ensure these Frampton, S. E., & Shillingsburg, M. A. (2019). Promot-
individuals make the social, language, and other ing the development of verbal responses using
instructive feedback. Journal of Applied Behavior Anal-
skill-based gains needed to reflect those of their ysis. Advance online publication. https://doi.org/10.
typically developing peers, additional research 1002/jaba.659
targeting specific instructional procedures that Gast, D. L., Doyle, P. M., Wolery, M., Ault, M. J., &
will increase the efficiency of learning for indi- Kolenda, J. L. (1994). Instructive feedback: Effects of
number and type. Journal of Behavioral Education, 4
viduals with ASD is necessary. (3), 313–334. https://doi.org/10.1007/BF01531985
Gilliam, J. E. (2013). Gilliam autism rating scale (3rd ed).
Pro-Ed.
REFERENCES Greer, R. D., & Ross, D. E. (2008). Verbal behavior anal-
ysis: Inducing and expanding new verbal capabilities in
Appelman, M., Vail, C. O., & Lieberman-Betz, R. G. children with language delays. Inc: Pearson Education.
(2014). The effects of constant time delay and Griffen, A. K., Schuster, J. W., & Morse, T. E. (1998).
instructive feedback on the acquisition of English The acquisition of instructive feedback: A comparison
and Spanish words. Journal of Early Intervention, of continuous versus intermittent presentation sched-
36(2), 131–148. https://doi.org/10.1177/105381511 ules. Education and Training in Mental Retardation
4563613 and Developmental Disabilities, 33(1), 42–61. http://
Caldwell, N. K., Wolery, M., Werts, M. G., & www.jstor.org/stable/23879041
Caldwell, Y. (1996). Embedding instructive feedback Haq, S. S., Zemantic, P. K., Kodak, T., LeBlanc, B., &
into teacher-student interactions during independent Ruppert, T. E. (2015). Examination of variables that
seat work. Journal of Behavioral Education, 6(4), affect the efficacy of instructive feedback. Behavioral
459–480. https://doi.org/10.1007/BF02110517 Interventions, 32(3), 206–216. https://doi.org/10.
Carroll, R. A., & Kodak, T. (2015). Using instructive 1002/bin.1470
feedback to increase response variability during Horne, P. J., & Lowe, C. F. (1996). On the origins of
intraverbal training for children with autism spectrum naming and other symbolic behavior. Journal of the
disorder. Analysis of Verbal Behavior, 31(2), 183–199. Experimental Analysis of Behavior, 65(1), 185–241.
https://doi.org/10.1007/s40616-015-0039-x https://doi.org/10.1901/jeab.1996.65-185
Cromer, K., Schuster, J. W., Collins, B. C., & Grisham- Leaf, J. B., Cihon, J. H., Alcalay, A., Mitchell, E.,
Brown, J. (1998). Teaching information on medical Townley-Cochran, D., Miller, K., … McEachin, J.
prescriptions using two instructive feedback sched- (2017). Instructive feedback embedded within group
ules. Journal of Behavioral Education, 8(1), 37–61. instruction for children diagnosed with autism spec-
https://doi.org/10.1023/A:1022812723663 trum disorder. Journal of Applied Behavior Analysis,
Dass, T. K., Kisamore, A. N., Vladescu, J. C., 50(2), 304–316. https://doi.org/10.1002/jaba.375
Reeve, K. F., Reeve, S. A., & Taylor-Santa, C. Miguel, C. F. (2016). Common and intraverbal bidirec-
(2018). Teaching children with autism spectrum dis- tional naming. The Analysis of Verbal Behavior, 32(2),
order to tact olfactory stimuli. Journal of Applied 125–138. https://doi.org/10.1007/s40616-016-
Behavior Analysis, 51(3), 538–552. https://doi.org/10. 0066-2
1002/jaba.470 Nottingham, C. L., Vladescu, J. C., & Kodak, T. M.
Delmolino, L., Hansford, A. P., Bamond, M. J., (2015). Incorporating additional targets into learning
Fiske, K. E., & LaRue, R. H. (2013). The use of trials for individuals with autism spectrum disorder.
instructive feedback for teaching language skills to Journal of Applied Behavior Analysis, 48(1), 227–232.
children with autism. Research in Autism Spectrum https://doi.org/10.1002/jaba.179
Disorders, 7(6), 648–661. https://doi.org/10.1016/j. Nottingham, C. L., Vladescu, J. C., Kodak, T., &
rasd.2013.02.015 Kisamore, A. N. (2017). Incorporating multiple sec-
Doyle, P. M., Schuster, J. W., & Meyer, S. (1996). ondary targets into learning trials for individuals with
Embedding extra stimuli in the task direction: Effects autism spectrum disorder. Journal of Applied Behavior
on learning of students with moderate mental retar- Analysis, 50(3), 653–661. https://doi.org/10.1002/
dation. The Journal of Special Education, 29(4), jaba.396
2302 Casey L. Nottingham et al.

Parrott, K. A., Schuster, J. W., Collins, B. C., & and Education, 50(2), 149–167. https://doi.org/10.
Gassaway, L. J. (2000). Simultaneous prompting and 1080/1034912032000089657
instructive feedback when teaching chained tasks. Tullis, C. A., Frampton, S. E., Delfs, C. H., &
Journal of Behavioral Education, 10(1), 3–19. http:// Shillingsburg, M. A. (2016). Teaching problem
doi.org/10.1023/A:1016639721684 explanations using instructive feedback. The Analysis
Reichow, B., & Wolery, M. (2011). Comparison of pro- of Verbal Behavior, 33(1), 64–79. https://doi.org/10.
gressive prompt delay with and without instructive 1007/s40616-016-0075-1
feedback. Journal of Applied Behavior Analysis, 44(2),
327–340. https://doi.org/10.1901/jaba.2011.44-327 Tullis, C. A., Gibbs, A. R., Butzer, M., & Hansen, S. G.
Ross, A. H., & Stevens, K. B. (2003). Teaching spelling (2019). A comparison of secondary target location in
of social studies content vocabulary prior to using the instructive feedback procedures. Advances in Neu-
vocabulary in inclusive learning environments: An rodevelopmental Disorders, 3(1), 45–53. https://doi.
examination of constant time delay, observational org/10.1007/s41252-018-0090-4
learning, and instructive feedback. Journal of Behav- Tullis, C. A., Marya, V., & Shillingsburg, M. A. (2019).
ioral Education, 12(4), 287–309. https://doi.org/10. Enhancing instruction via instructive feedback for a
1023/A:1025917824403 child with autism using a speech-generating device.
Schnell, L. K., Vladescu, J. C., Kodak, T., & Analysis of Verbal Behavior, 35(1), 103–112. https://
Nottingham, C. L. (2018). Comparing procedures doi.org/10.1007/s40616-018-0096-z
on the acquisition and generalization of tacts for chil- Vladescu, J. C., & Kodak, T. M. (2013). Increasing
dren with autism spectrum disorder. Journal of instructional efficiency by presenting additional stim-
Applied Behavior Analysis, 51(4), 769–783. https:// uli in learning trials for children with autism spec-
doi.org/10.1002/jaba.480 trum disorders. Journal of Applied Behavior Analysis,
Sindelar, P. T., Rosenberg, M. S., & Wilson, R. J. 46(4), 805–816. https://doi.org/10.1002/jaba.70
(1985). An adapted alternating treatments design for Werts, M. G., Hoffman, E. M., & Darcy, C. (2011).
instructional research. Education and Treatment of Acquisition of instructive feedback: Relation to target
Children, 8(1), 67–76. https://www.jstor.org/stable/ stimulus. Education and Training in Autism and
42898888 Developmental Disabilities, 46(1), 134–149. https://
Sundberg, M. L. (2008). Verbal behavior milestones assess- www.jstor.org/stable/23880037
ment and placement program: The VB-MAPP. AVB
Press. Wolery, M., Gast, D. L., & Ledford, J. R. (2014). Com-
Sundberg, M. L. (2016). Verbal stimulus control and the parison designs. In D. L. Gast & J. R. Ledford
intraverbal relation. Analysis of Verbal Behavior, 32 (Eds.), Single case research methodology: Applications in
(2), 107–124. https://doi.org/10.1007/s40616-016- special education and behavioral sciences
0065-3 (pp. 297–345). Routledge.
Tekin-Iftar, E., Acar, G., & Kurt, O. (2003). The effects
of simultaneous prompting on teaching expressive Received May 9, 2019
identification of objects: An instructive feedback Final acceptance February 10, 2020
study. International Journal of Disability, Development Action Editor, M. Alice Shillingsburg

You might also like