Download as pdf or txt
Download as pdf or txt
You are on page 1of 14

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2017, 9999, n/a–n/a NUMBER 9999 ()

TEACHING MANDS FOR INFORMATION USING ‘WHEN’


TO CHILDREN WITH AUTISM
ROBIN K. LANDA
MARCUS AUTISM CENTER

BETHANY HANSEN
MARCUS AUTISM CENTER AND EMORY UNIVERSITY SCHOOL OF MEDICINE

AND

M. ALICE SHILLINGSBURG
MARCUS AUTISM CENTER AND EMORY UNIVERSITY SCHOOL OF MEDICINE

Previous research has evaluated contrived motivating operations to teach mands for information.
However, literature evaluating acquisition of the mand when? is comparatively limited. As an
extension of Shillingsburg, Bowen, Valentino, & Pierce (2014), we taught three children with
autism to engage in mands for information using when under alternating conditions in which a
contrived establishing operation was present (EOP) or absent (EOA). Following training with a
constant prompt delay, all participants acquired the mand for information and demonstrated
correct use of the provided information and a decrease in inappropriate attempts to access
restricted items.
Key words: autism, establishing operations, mands for information, requesting when, verbal
behavior

Mands for information occur under the con- behavior contacts reinforcement. In other
trol of an establishing operation (EO) and words, asking for the location of the shoe is
result in information that often evokes addi- likely more efficient than searching the room.
tional behavior maintained by a related rein- Several studies illustrate the utility of contriv-
forcer. For example, when it is time to go ing establishing operations (EOs) to teach chil-
outside to play a child may ask, “Where is my dren to mand for information, often in the form
shoe?” to which an adult might respond, “It’s of wh— questions (e.g., Betz, Higbee, & Pol-
by the door.” In this example, the missing shoe lard, 2010; Endicott & Higbee, 2007; Lechago,
functions as an EO that increases the value of Carr, Grow, Love, & Almason, 2010; Shillings-
the information related to its location. The burg & Valentino, 2011; Shillingsburg, Valen-
information, “It’s by the door,” both reinforces tino, Bowen, Bradley, & Zavatkay, 2011;
the mand and evokes a chain of behavior Sundberg, Loeb, Hale, & Eigenheer, 2002;
(i.e., going to indicated location) that ends with Williams, Donley, & Keller, 2000). Although
access to the shoe. Mands for information may contrived EOs may increase the value of infor-
increase the efficiency with which children’s mation, analysis of discriminated responding
through EO manipulation is necessary to deter-
Correspondence should be sent to Alice Shillingsburg, mine whether the acquired skill functions as a
PhD, BCBA-D, Program Director, Language and Learning mand. Recent studies have compared question-
Clinic, Marcus Autism Center, 1920 Briarcliff Rd. NE, asking under alternating conditions in which a
Atlanta, GA 30329. E-mail: alice.shillingsburg@choa.org_
404-785-9400. contrived EO was present (EOP) or absent
doi: 10.1002/jaba.387 (EOA) to verify functional control for the
© 2017 Society for the Experimental Analysis of Behavior
1
2 ROBIN K. LANDA et al.

mands where? (Howlett, Sidener, Progar, & treatment. In addition, the authors reported
Sidener, 2011), how? (Lechago, Howell, Cacca- participant approaches to the correct person in
velle, & Peterson, 2013; Shillingsburg, possession of the item (for who?) and location
Bowen, & Valentino, 2014), which? and who? in which the item was stored (for which?). Fol-
(Shillingsburg, Bowen, Valentino, & Pierce, lowing training, all participants emitted mands
2014), and I don’t know, please tell me for information under EOP conditions. In
(Ingvarsson & Hollobaugh, 2010), which may addition, participants demonstrated an increase
be considered a form of what? For example, in correct approach responses following acquisi-
Howlett et al. (2011) taught two children with tion of mands for information in the EOP con-
language delays to mand for information related dition, indicating correct use of the provided
to missing items, using where under alternating information. Results indicate that the acquired
EOP and EOA conditions. Participants were response not only functioned as a mand, but
instructed to select and retrieve a preferred item also increased the likelihood that children’s
with which to play at the start of each trial. Dur- attempts to access reinforcers (i.e., approach
ing EOA trials, the preferred item was present responses) would be successful.
in its typical location, whereas during EOP In comparison to other mands for informa-
trials, it was missing. Script fading during EOP tion, research investigating the mand when? is
trials (both participants) and error correction limited (Shillingsburg et al., 2011). The mand
during EOA trials (one participant) were used when? may be evoked by a variety of verbal and
to teach children to mand where? Following visual stimuli correlated with delayed access to
treatment, both participants emitted the ques- reinforcers. For example, following a child’s
tion where? exclusively during the EOP condi- request to play outside, parents might respond
tion. The inclusion of participants’ initial “not right now,” which may evoke the mand
selection of the tangible reinforcer at the begin- “when?” Alternatively, children may ask,
ning of trials increased the likelihood of EO “When may I play outside?” following an
control, and the analysis of discriminated observation that it is raining. Typical responses
responding under the EOP condition, relative to the mand when? may include contingency-
to the EOA condition, verified it. specifying information, such as “after you do
Analyzing discriminated responding under your homework”; time-specifying information,
EOP and EOA conditions allows for verifica- such as “in 15 minutes”; or event-specifying
tion of control by the relevant EO. However, information, such as “after I finish cooking” or;
practical analysis requires additional measures, “after it stops raining”. Information may also
such as the extent to which participants be provided that does not enable accurate pre-
respond to information delivered contingent on diction of reinforcer availability, such as “I
the mand. Shillingsburg, Bowen, Valentino, & don’t know” or “later”. Teaching children to
Pierce (2014) taught the mands which? and obtain and respond to information related to
who? to children with autism under contrived reinforcer availability may allow children to
conditions in which information was needed access reinforcers in an efficient manner during
(EOP) or not needed (EOA) to access a pre- delays when emission of problem behaviors is
ferred item. To enhance assurance of EO con- likely occur. For example, the response—“after
trol, trials were conducted following you do your homework”—may evoke a chain
participant-initiated requests for items. A con- of responses that exerts discriminative control
stant time-delay (Schuster, Gast, Wolery, & over mands. In other words, children may learn
Guiltinan, 1988) procedure was used to to emit a subsequent mand following, but not
prompt the mand for information during prior to, completion of homework. This may
MANDS FOR INFORMATION USING ‘WHEN’ 3

increase both the proportion of mands that designated task and discriminative control of
result in reinforcement and the probability of requests. Evaluating the use of contingency-
appropriate behavior during delays. specifying information might be an important
Despite the potential benefits of teaching prerequisite to teaching a complex repertoire of
children to mand “when?” we are aware of only responding to information related to reinforcer
one study investigating procedures to teach this delays. Therefore, the primary purpose of this
skill (Shillingsburg et al., 2011). Shillingsburg study was to extend Shillingsburg, Bowen,
et al. taught two children with autism to Valentino, & Pierce (2014) by teaching the
request information using who, which, where, mand for information when? under alternating
and when questions. To contrive the relevant EOP and EOA conditions and by evaluating
EO for the question when? therapists presented the utility of the mand by examining partici-
a verbal stimulus such as, “We are going to play pant use of contingency-specifying information.
on the computer, but not right now.” During
treatment, if participants did not emit a mand
METHOD
within 3 s, the verbal stimulus was represented
with a vocal prompt. Information related to a Participants and Setting
behavior requirement that led to the targeted Three children diagnosed with autism and
activity was provided following prompted and enrolled in an intensive behavior intervention
independent mands for information. For exam- program focused on communication deficits
ple, if the participant asked, “When can I participated in the study. Each emitted
play?” the therapist responded with, “after you multiple-word vocal mands, could complete a
give me a high five.” Access to the computer variety of independent academic and self-care
was provided following compliance with the tasks, and consistently followed receptive
behavior requirement. Both participants instructions. Haley, a 6-year-old female, exhib-
acquired the mand when? However, several lim- ited no mands for information of any type.
itations warrant additional research. First, this Brian, a 7-year-old male, exhibited the mands
study did not include a preference assessment for information who? and which? but did not
prior to trials and therapists did not wait for an exhibit the mand when? Daisy, a 6-year-old
independent mand for the activity to initiate female, was observed to emit mands for infor-
trials. Doing either of these would have indi- mation, including what? and who? but did not
cated that the delayed activities were preferred exhibit the mand when? Both Brian and Daisy
at the time, enhancing assurance of EO con- had been previously taught to mand using wh-
trol. Second, the authors did not conduct alter- questions using similar EOA and EOP proce-
nating EOP and EOA conditions to verify dures. All participants displayed a history of
functional control as a mand. Third, this study intolerance to denied or delayed reinforcers in
did not include data representing use of the the form of persistent mands, repeated attempts
information provided to demonstrate utility of to grab items, tantrums, or some combination
the skill. In other words, it is not clear if the of these behaviors. The Verbal Behavior Mile-
participant “used” the information to complete stones Assessment and Placement Program
a task or wait an appropriate interval until the (VB-MAPP; Sundberg, 2008) was conducted
activity was available. for each participant upon admission into the
Use of contingency-specifying information clinical program. At the time of participation,
might be evaluated by examining the extent to participants’ overall VB-MAPP scores ranged
which the information (e.g., “after you do your from 96–125. Scores fell primarily in the
homework”) results in completion of the beginning Level 3 range for the Mand (range,
4 ROBIN K. LANDA et al.

10–12), Tact (range, 9.5-11), and Listener with no more than 5 consecutive seconds of
(range, 10–11) domains, corresponding with a off-task behavior. We considered the child to
developmental age of 30–48 months. Scores for be off-task when he or she physically oriented
the Intraverbal domain fell primarily within the away from materials (e.g., moved out of seat or
Level 2 range (range, 3.5-5.5), corresponding away from the designated area) or when the
with a developmental age of 18–30 months child shifted his or her gaze away from materi-
(individual scores available upon request). als in the absence of manipulating them. That
Sessions were conducted within the partici- is, shifting eye gaze while drying hands would
pants’ respective classrooms, which contained not have been considered off-task behavior, but
several desks and chairs, a large table for group shifting eye gaze while standing at the sink
instruction, play areas, and leisure and aca- without touching the faucet, soap, or paper
demic materials. During each session, partici- towels would be considered off-task. Initiations
pants were instructed to complete various were defined as vocal or physical attempts to
behavior requirements, such as cleaning a table, access preferred items (i.e., by manding, reach-
picking up toys, washing hands, sorting silver- ing, or grabbing).
ware, and common classroom tasks, such as Trained observers used paper and pencil to
independent activities and cleaning up beanbag record data across 5-trial (preassessment) or 10-
squares used for seating arrangements. Materi- trial (mands for information) sessions. Mands
als necessary to complete these activities were for information were scored as independent
provided at every session and included toys, when the participant emitted the vocal response
independent activities (e.g., matching games within 5 s of a therapist’s denial of the item
and worksheets), soap, towels, sanitizing wipes, and were scored as prompted when they
plastic utensils, a silverware tray (Haley only), occurred within 5 s of a therapist’s prompt.
and beanbags (Daisy only). Although the time Correct completion of the behavior requirement
required to complete each task differed, in gen- was scored as independent when the participant
eral, each task took no longer than 2 min to initiated the therapist-instructed behavior
complete. Highly preferred items were visually within 5 s of the contingency-specifying infor-
present, but out of reach of the participant dur- mation. The completion of the behavior
ing sessions. requirement was scored as prompted when the
participant initiated the behavior within 5 s of a
therapist’s prompt. Initiations for reinforcers
Response Measurement and Interobserver were scored as appropriate when they occurred
Agreement (IOA) within 10 s of completion of the therapist-
The primary dependent variable was the per- instructed behavior (e.g., after washing hands)
centage of trials with mands for information and as inappropriate when they occurred prior
during EOP and EOA conditions. Mands for to completion of the behavior. The number of
information were defined as the vocal response, trials with independent responses, prompted
“When?” (Brian and Daisy) or “When can I responses, and nonresponses for participants’
have [item]?” (Haley). Secondary measures mands for information and correct performance
included the percentage of trials with correct of the behavior requirement was divided by the
performance of the behavior requirement and total number of trials to yield a percentage of
the percentage of appropriate and inappropriate trials with each type of response. The frequency
initiations for reinforcers. Correct performance of appropriate or inappropriate initiations was
of the behavior requirement was defined as divided by the total frequency of all initiations
completing the therapist-instructed behavior to yield percentages of each.
MANDS FOR INFORMATION USING ‘WHEN’ 5

Interobserver agreement (IOA) was assessed statements prior to mands for information train-
by having a second, independent observer ing if they did not do so in the preassessment.
simultaneously collect data during 35% and Sessions consisted of five trials and were com-
100% of preassessment sessions for Haley and pleted within a few days. We conducted sessions
Daisy, respectively. IOA for preassessment ses- during breaks from academic activities and class-
sions was not collected for Brian. During mands room routines, during which we presented a
for information sessions, agreement data were variety of preferred items in view but out of
collected during 50% of sessions for Haley and reach. Preferred items included previously iden-
Daisy and 46% of sessions for Brian. Mean tified edible and tangible reinforcers (e.g., juice,
agreement was calculated using the total agree- chips, books, iPad). If the participant manded
ment method. During the preassessment, mean for an item, we either provided immediate
agreement was 95% (range, 90%–100%) for access to the reinforcer (about one third of the
Haley and 98% (range, 96%–100%) for Daisy. time) or initiated a trial. If the participant did
For mands for information sessions, mean not emit a mand, we provided 30-s access to
agreement was 96% (range, 90%–100%) for each item and then removed it until a mand was
Haley, 97% (range, 90%–100%) for Brian, and evoked. The therapist initiated trials by
96% for Daisy (range, 85%–100%). Procedural responding to a mand such as, “May I have a
integrity data were collected on whether the snack?” with a contingency-specifying statement
therapist initiated trials following a participant’s such as, “After you put away your toys,” and
mand for an item, responded to mands accord- starting a timer. Behavior requirements were
ing to the designated condition, differentially selected based on parent and teacher reports of
reinforced appropriate initiations, and employed common behavioral expectations during delays.
prompting as outlined by the procedures. Data For example, we selected sorting utensils as one
were collected for at least 45% of all sessions for of Haley’s behavior requirements, as Haley’s
all participants and indicated 100% integrity. mother expressed she would like her to com-
plete household chores, such as putting the
dishes away, prior to watching television or
Preassessment playing on the computer. All items necessary to
Preassessments were conducted prior to eval- complete the behavior requirements were pres-
uating mands for information. The purpose of ent across all trials and any potential extraneous
the preassessment was to evaluate if prompts or visual cues were withheld. Behavior
contingency-specifying information given by an requirements were randomly assigned across
adult resulted in completion of the required trials. We refer to repeated item or activity
behavior followed by appropriate initiations for mands that occurred during the trial as initia-
the item (i.e., manding for the item following, tions, to distinguish them from the mand that
but not prior to, completion of the behavior started a trial. Inappropriate initiations
requirement). These prerequisite behaviors (i.e., repeated attempts to access the denied item
would be required to access the initially denied prior to completing the behavior requirement)
preferred item during mands for information were ignored or blocked. Appropriate initiations
training. We therefore assumed that (i.e., attempts to access the item within 10 s of
contingency-specifying information would not completing the required behavior) resulted in
act as an effective consequence for mands for 30 s access to the reinforcer. Trials ended within
information during training if it did not evoke 10 s of completion of the behavior requirement
these prerequisites. Therefore, we taught partici- or following 45 s, if the behavior requirement
pants to respond to contingency-specifying was not initiated.
6 ROBIN K. LANDA et al.

Mastery criteria for the preassessment con- Session procedures were identical across all con-
sisted of 100% correct behavior performance ditions with the exception that mands for infor-
and 100% appropriate initiations across one mation were never prompted during baseline
(Daisy) or two (Haley and Brian) consecutive and posttraining sessions. Sessions generally
sessions. Participant specific criteria were cho- lasted 5–10 min, were implemented during
sen based on previous experience with the par- breaks between academic activities and class-
ticipants’ learning histories in order to keep the room routines, and consisted of 10 trials with
preassessment procedures as short as possible. random interspersal of five EOP and five EOA
Once mastery was met, the participant pro- trials. For Haley and Brian, access to preferred
ceeded with baseline and intervention condi- items and activities used during sessions was
tions for teaching mands for information. If restricted for approximately 10 min prior to
these criteria were not met, training was con- and following sessions. For Daisy, access to pre-
ducted prior to moving forward. During train- ferred items and activities used during sessions
ing, a constant time delay was implemented to was restricted outside of sessions due to a his-
teach the participant to complete the behavior tory of rapid satiation of these highest preferred
requirement and to subsequently mand for the items.
reinforcer. Using combined vocal and gestural We used procedures described in the preas-
prompts, the therapist immediately prompted sessment to evoke a mand for an item or activ-
the participant to complete the behavior ity prior to each trial. As in the preassessment,
requirement within 0–1 s following the mands for items or activities resulted in either
contingency-specific statement. A vocal prompt immediate reinforcement (about one third of
was subsequently delivered within 0–1 s of the time) or initiation of a trial. During EOP
completion of the behavior to prompt the par- trials, the therapist responded to mands for
ticipant to vocally mand for the reinforcer items or activities with a denial statement
(i.e., emit an appropriate initiation). Following (e.g., “not right now”). During EOA trials, the
one session with 100% of trials with the correct therapist responded to mands with a denial
behavior performed and 100% appropriate statement and a contingency-specifying state-
initiations, the prompt delay was extended to ment (e.g., “not right now, after you wash your
3 s. Training continued until criteria for mov- hands”). Therapists withheld the denied item
ing on to baseline for mands for information or activity during all trials, until they completed
were met. All three participants met mastery one of five preselected behavior requirements
criteria on the preassessment (data not shown (e.g., washing hands) and engaged in an appro-
but available upon request); thus, they were priate initiation. The only way participants
able to complete all behaviors in the could access the denied reinforcer during EOP
contingency-specific statement used in the trials was by either “guessing” the correct
study. Brian and Daisy met mastery in baseline behavior requirement, which was unlikely, or
and did not require training. Haley’s respond- emitting the mand for information when,
ing was between 60%–80% in baseline and, which resulted in delivery of the contingency-
therefore, training was conducted for 10 sessions specifying statement. We presumed that with-
until mastery was met. holding contingency-specifying statements at
the beginning of the EOP trial and presenting
them only following mands for information
Mands for Information would evoke and reinforce mands for informa-
We evaluated mands for information across tion, respectively. By contrast, we presumed
baseline, training, and posttraining sessions. that presenting the contingency-specifying
MANDS FOR INFORMATION USING ‘WHEN’ 7

statement at the start of each EOA trial would initiations (i.e. initiations following task com-
render the EO for mands for information pletion) were reinforced. Specifically, if a partic-
absent. Denial statements included the phrases ipant “guessed” the behavioral requirement and
not right now, you can have that later, and that’s completed it and then subsequently manded
unavailable. Contingency-specifying statements again for the item, it was delivered. We with-
included those assessed in the preassessment. held prompts to complete the target behavior
Denial and contingency-specifying statements requirement and emit appropriate initiations
were randomly assigned within conditions prior across all phases to allow for an assessment of
to the start of the session. That is, all five whether the relevant EO was in place. In other
behavior requirements were targeted during words, lack of completing the behavior require-
both conditions, but the order in which they ment or subsequent initiation for the reinforcer
were targeted varied. The order of EOA and may indicate that the EO relevant to the termi-
EOP trials was randomly determined prior to nal reinforcer was absent; prompts were with-
the session. held to ensure that when these behaviors did
Following each denial or contingency- occur, they were under control of the relevant
specifying statement, the therapist started a EO, rather than discriminative control of sti-
timer and allowed 5 s for the participant to muli correlated with prompts.
emit a mand for information or to initiate the Training sessions were identical to baseline
behavior requirement. The therapist reinforced and posttraining sessions except that we used a
all mands for information by presenting the constant time delay to teach the participants to
contingency-specifying statement (e.g., “after mand for information during EOP trials. With a
you pick up your toys”). Inappropriate initia- 0- to 1-s delay, immediately following the denial
tions including vocal mands for the preferred statement on EOP trials, the therapist delivered a
item and physical attempts to access the item textual (Haley) or echoic (Brian and Daisy)
were ignored. It should be noted, however, that prompt to emit the mand for information. Once
none of the participants engaged in physical the participant emitted a correct mand for infor-
attempts. If they had, the physical attempt mation across 100% of EOP trials with the 0- to
would have been blocked. All appropriate 1-s delay for one session, the prompt delay was
initiations (i.e., initiations following the com- extended to 3 s to allow the opportunity for an
pletion of the behavior requirement) were rein- independent response. If an independent correct
forced with the manded item. Trials ended response was not emitted within 3 s of the EO
either immediately after an appropriate initia- presentation, a prompt was provided. No pro-
tion was emitted or following 10 s after the grammed consequences were in place for inap-
behavior requirement was completed, which- propriate initiations or errors that occurred prior
ever came first. That is, if the participant com- to this prompt. Prompts were removed and post-
pleted the behavior requirement, but did not training was conducted when participants
emit an appropriate initiation by manding emitted a correct mand for information across
again for the item within 10 s, the trial was ter- 100% of EOP trials and no more than 20% of
minated. Therapists then presented two to EOA trials for two consecutive sessions at the 3-s
three mastered tasks and reinforced compliance prompt delay.
with praise and a break, rather than the tangi-
ble reinforcer, to contrive the relevant EO for
the next trial. If the participant performed the RESULTS
correct targeted behavior on EOP trials without During baseline, none of the participants
manding for information, appropriate emitted mands for information during EOP
8 ROBIN K. LANDA et al.

trials (Figure 1). Throughout the study, all par- was given (Figure 2). By contrast, during EOP
ticipants continued to complete behavior trials, when the contingency-specifying infor-
requirements during 80%-100% of EOA trials mation was withheld, all of the participants
in which the contingency-specifying statement were unable to consistently perform the correct

Baseline Training Posttraining


100

80

60

40

20
Haley
Mands for Information - When (% of Trials)

0
100

80

60 EOP (I)
EOP (P)
EOA
40

20
Brian

0
100

80

60

40

20
Daisy

0
5 10 15 20
Sessions

Figure 1. Percentage of trials with independent mands for information (I) under EOA and EOP conditions and
prompted mands for information (P) under EOP condition only, during baseline, training, and posttraining.
MANDS FOR INFORMATION USING ‘WHEN’ 9

Baseline Training Posttraining


100

80

60

40

20
Haley

0
Correct Behavior Performed (% of Trials)

100

80

60
EOP
EOA
40

20
Brian

0
100

80

60

40

20
Daisy

0
5 10 15 20
Sessions

Figure 2. Percentage of trials in which participants correctly completed the behavior requirement to access reinfor-
cers under EOA and EOP conditions during baseline, training, and posttraining.

behavior requirement (Figure 2) and displayed participant who completed tasks during the
persistent mands for the denied item, as indi- EOP trials, resulting in a few correct “guesses.”
cated by high levels of inappropriate reinforcer Daisy’s completion of tasks when denied access
initiations (Figure 3). Daisy was the only to preferred items was presumably due to a
10 ROBIN K. LANDA et al.

Baseline Training Posttraining


100

80

60

40

20
Haley

0
100
Percentage of Initiations

80

60
Appropriate
40 Inappropriate

20
Brian

0
100

80

60

40

20
Daisy

0
5 10 15 20
Sessions

Figure 3. Percentage of appropriate and inappropriate initiations for reinforcers under EOP conditions during base-
line, training, and posttraining.

history of redirection to similar tasks following being told which task to complete and subse-
denials. Given that the behavior requirements quently emitted an appropriate initiation for
were randomized before each session, she occa- the denied item. This is reflected in the few ses-
sionally completed the correct task without sions with low correct responding to the
MANDS FOR INFORMATION USING ‘WHEN’ 11

behavior requirements (Figure 2). Data on response. The second is that absence of
initiations in EOA trials are not displayed, as contingency-specifying information during
there was no change in initiations in the EOA EOP conditions served as an establishing opera-
condition following introduction of training. tion resulting in acquisition of a mand, whereas
Participants never (Haley) or rarely (Daisy and the presence of contingency-specifying informa-
Brian) emitted inappropriate initiations and tion during EOA conditions served as an abol-
almost always emitted appropriate initiations ishing operation.
during EOA trials across all phases (data availa- Several factors increase confidence in the
ble from the corresponding author). interpretation of EO control. First, all trials
During mands for information training, all began following a participant’s mand for a
participants began responding to prompted highly preferred item, increasing the likelihood
trials to emit the mand when? during EOP that an EO relevant to the terminal reinforcer
trials (Figure 1). Prompts were delayed to 3 s was present at the beginning of each trial. Sec-
following the first session, for all participants. ond, all participants demonstrated a general
Haley met mastery criteria in 10 sessions, Brian degree of consistency in completion of the
met mastery in 3 sessions, and Daisy met mas- behavior requirement and subsequent mand for
tery in 11 sessions. All participants emitted the the item, across both EOP and EOA condi-
mand for information when? under EOP trials tions, increasing confidence that the EO rele-
exclusively, indicating the mand was under vant to the terminal reinforcer was present
control of the relevant establishing operation. throughout each trial. Third, there were no
Acquisition of the mand for information programmed differential consequences for ask-
resulted in an increase in the correct behavior ing “when?” across either condition (i.e., asking
performed and an increase in the proportion of “when?” always resulted in delivery of
appropriate initiations relative to inappropriate contingency-specifying information). However,
initiations during EOP trials for all participants. only Daisy emitted the when response during
Similar results were observed in the absence of EOA conditions and contacted this
all prompts during the posttraining condition. contingency.
Participants’ use of information was demon-
strated through completion of the behavior
DISCUSSION requirement and a high proportion of appropri-
The results of this study extend prior ate initiations (i.e., discriminated mands for
research by demonstrating successful methods items), highlighting the potential benefits of
for teaching the mand for information when? teaching the mand when? In baseline, partici-
under EOP and EOA conditions and add to pants were able to complete alternative activ-
the growing body of literature reporting func- ities prior to attempting to access reinforcers
tional use of information acquired via newly when provided with denials and specific
taught mands. Following training, the mand instructions (i.e., during EOA conditions in
when? occurred exclusively when information which adults presented statements such as “not
was withheld (EOP). These results may be right now, after you …”), but not when adults
interpreted in two ways. The first is that the simply denied reinforcers with no additional
therapist’s verbal response during EOP condi- direction (i.e., during EOP conditions in which
tions served as a discriminative stimulus, adults presented statements such as “not right
whereas the therapist’s verbal response during now”). It should be noted that Haley and Brian
EOA conditions served as a neutral stimulus, emitted continuous inappropriate initiations for
resulting in acquisition of a discriminated reinforcers following denials and never
12 ROBIN K. LANDA et al.

attempted to engage in any of the targeted nor conducted a functional analysis of more
behavior requirements. However, acquisition of severe problem behavior. Future research to
the mand when? resulted in a decrease in inap- identify the conditions for teaching this skill is
propriate initiations and an increase in correct thus warranted, especially with children who
completion of behavior requirements. engage in more severe problem behavior follow-
Participants’ caregivers reported that inap- ing denials, as well as those with fewer prereq-
propriate initiations were problematic and typi- uisite skills than those included in the current
cally preceded more severe forms of problem study. The finding that acquisition of the mand
behavior (e.g., crying, screaming, flopping, when? resulted in a decrease in repetitive mands
aggression, or property destruction). We and an increase in appropriate alternative
observed high levels of inappropriate initiations behavior may have relevance to other contexts
exclusively in the EOP condition during base- in which reinforcers are not immediately availa-
line, indicating that they were evoked by rein- ble contingent on appropriate behaviors. For
forcer denials in the absence of additional instance, caregivers may have a variety of expec-
directives. Reinforcement delays or denials are tations for what children should do following
often unavoidable outside of highly controlled reinforcer denials. Children may be expected to
settings and, in the absence of treatment, may complete alternative tasks (e.g., “after you do
result in the emergence of inappropriate behav- your homework”) or engage in an alternative
ior, including high-rate mands (e.g., Hanley, activity without disruptive behavior for a speci-
Iwata, & Thompson, 2001). Treatments that fied amount of time (e.g., “in 5 minutes”),
have been successful in increasing tolerance to until an event elapses (e.g., “when it stops rain-
denials or delays often require specific adult ing outside”), or for an unpredictable duration
responses, such as delivery of instructions and (e.g., “I’m not sure when that will be availa-
differential reinforcement of compliance ble”). Therefore, researchers should also iden-
(e.g., contingency-based delay training; Hanley, tify the utility of teaching children to both
Jin, Vanselow, & Hanratty, 2014), provision of respond to and recruit information regarding a
alternative reinforcers (e.g., Austin & Tiger, variety of caregiver expectations following rein-
2015), or arrangement of discriminative stimuli forcer denials. Of particular interest may be the
(e.g., multiple schedules with contrived stimuli; duration of the delay specified in the informa-
Betz, Fisher, Roane, Mintz, & Owen, 2013) tion given by the caregiver.
prior to or following denial cues. However, it Some limitations to the current study war-
seems reasonable to expect that caregivers who rant additional research. First, we investigated
are busy, are distracted, or have not received only the use of contingency-specifying informa-
specialized training with these interventions tion. To increase the functionality and mainte-
may simply deny reinforcers without providing nance of the mand when? future research
additional instructions, alternative reinforcers, should investigate methods for teaching chil-
or other discriminative stimuli. Teaching chil- dren to respond appropriately to a variety of
dren to mand for information regarding what types of information related to reinforcer availa-
to do following such reinforcer denials may bility. For example, asking, “When is dinner?”
have relevance to the prevention and treatment or, “When can I go to the park?” does not nec-
of problem behavior, given that teaching chil- essarily lead to access to or shorten the wait
dren to mand when? remediated inappropriate time to those events. Understanding the multi-
initiations for all participants in the current ple reinforcers responsible for use of the mand
study. However, we neither evaluated the func- when? and the common expectations during
tional reinforcer for inappropriate initiations delays or denials is important to developing
MANDS FOR INFORMATION USING ‘WHEN’ 13

effective methods of teaching this skill and pro- Hanley, G. P., Jin, C. S., Vanselow, N. R., &
moting maintenance and generalization of its Hanratty, L. A. (2014). Producing meaningful
improvements in problem behavior of children with
use. Second, participants were taught to emit autism via synthesized analyses and treatments. Jour-
the mand when? following a therapist’s denial nal of Applied Behavior Analysis, 47, 16–36. https://
response. Because initiating trials following a doi.org/10.1002/jaba.106
Howlett, M. A., Sidener, T. M., Progar, P. R., &
participant’s initial mand for an item may Sidener, D. W. (2011). Manipulation of motivating
increase assurance of EO control, conducting operations and the use of a script-fading procedure to
initial teaching in this context might be desira- teach mands for location to children with language
ble. However, there are other occasions in delays. Journal of Applied Behavior Analysis, 44,
943–947. https://doi.org/10.1901/jaba.2011.44-943
which it is more socially appropriate to emit Ingvarsson, E. T., & Hollobaugh, T. (2010). Acquisition
the mand when? in the context of naturally of intraverbal behavior: Teaching children with
occurring stimuli that signal the unavailability autism to mand for answers to questions. Journal of
Applied Behavior Analysis, 43, 1–17. https://doi.org/
of reinforcers, such as when another child is 10.1901/jaba.2010.43-1
playing with a preferred item or when care- Lechago, S. A., Carr, J. E., Grow, L. L., Love, J. R., &
givers are engaged in busy activities and not Almason, S. M. (2010). Mands for information gen-
necessarily when a denial is given. Additional eralize across establishing operations. Journal of
Applied Behavior Analysis, 43, 381–395. https://doi.
research evaluating methods for teaching mands org/10.1901/jaba.2010.43-381
for when under such conditions seems war- Lechago, S. A., Howell, A., Caccavale, M. N., & Peterson,
ranted. Other researchers may also consider C. W. (2013). Teaching “how?” mand-for-information
evaluating similar procedures for teaching other frames to children with autism. Journal of Applied
Behavior Analysis, 46, 781–791. doi: 10.1002/jaba.71
mands for information such as why? Schuster, J. W., Gast, D. L., Wolery, M., & Guiltinan, S.
(1988). The effectiveness of a constant time-delay
procedure to teach chained responses to adolescents
REFERENCES with mental retardation. Journal of Applied Behavior
Analysis, 21, 169–178. https://doi.org/10.1901/jaba.
Austin, J. E., & Tiger, J. H. (2015). Providing alternative 1988.21-169
reinforcers to facilitate tolerance to delayed reinforce- Shillingsburg, M. A., Bowen, C. N., & Valentino, A. L.
ment following functional communication training. (2014). Mands for information using “how” under
Journal of Applied Behavior Analysis, 48, 663–668. EO-absent and EO-present conditions. The Analysis
https://doi.org/10.1002/jaba.215 of Verbal Behavior, 30, 54–61. https://doi.org/10.
Betz, A. M., Fisher, W. W., Roane, H. S., 1007/s40616-013-0002-7
Mintz, J. C., & Owen, T. M. (2013). A component Shillingsburg, M. A., Bowen, C. N., Valentino, A. L., &
analysis of schedule thinning during functional com- Pierce, L. E. (2014). Mands for information using
munication training. Journal of Applied Behavior “who?” and “which?” in the presence of establishing
Analysis, 46, 219–241. https://doi.org/10.1002/ and abolishing operations. Journal of Applied Behavior
jaba.23 Analysis, 47, 136–150. https://doi.org/10.1002/
Betz, A. M., Higbee, T. S., & Pollard, J. S. (2010). Pro- jaba.101
moting generalization of mands for information used Shillingsburg, M. A., & Valentino, A. L. (2011). Teach-
by young children with autism. Research in Autism ing a child with autism to mand for information
Spectrum Disorders, 4, 501–508. https://doi.org/10. using “how.” The Analysis of Verbal Behavior, 27,
1016/j.rasd.2009.11.007 179–184.
Endicott, K., & Higbee, T. S. (2007). Contriving moti- Shillingsburg, M. A., Valentino, A. L., Bowen, C. N.,
vating operations to evoke mands for information in Bradley, D., & Zavatkay, D. (2011). Teaching chil-
preschoolers with autism. Research in Autism Spectrum dren with autism to request information. Research in
Disorders, 1, 210–217. https://doi.org/10.1016/j.rasd. Autism Spectrum Disorders, 5, 670–679. https://doi.
2006.10.003 org/10.1016/j.rasd.2010.08.004
Hanley, G. P., Iwata, B. A., & Thompson, R. H. (2001). Sundberg, M. L. (2008). Verbal behavior milestones assess-
Reinforcement schedule thinning following treatment ment and placement program: The VB-MAPP. Con-
with functional communication training. Journal of cord, CA: AVB Press.
Applied Behavior Analysis, 34, 17–38. https://doi.org/ Sundberg, M. L., Loeb, M., Hale, L., & Eigenheer, P.
10.1901/jaba.2001.34-17 (2002). Contriving establishing operations to teach
14 ROBIN K. LANDA et al.

mands for information. The Analysis of Verbal Behav- Received January 26, 2016
ior, 18, 15–29. Final acceptance July 7, 2016
Williams, G., Donley, C. R., & Keller, J. W. (2000). Action Editor, Alison Betz
Teaching children with autism to ask questions about
hidden objects. Journal of Applied Behavior Analysis,
33, 627–630. https://doi.org/10.1901/jaba.2000.
33-627.

You might also like