You are on page 1of 17

The Analysis of Verbal Behavior 2008, 24, 6985

Discrete Trial Instruction vs. Mand Training for Teaching

Children With Autism to Make Requests

Heather K. Jennett, Kennedy Krieger Institute

Sandra L. Harris and Lara Delmolino
Rutgers, The State University of New Jersey

The present study compared the effects of discrete trial instruction (DTI) and mand training on the
acquisition of independent requests in 6 children with autism. Two multiple-probe designs across partici-
pants were conducted with 3 participants receiving mand training followed by DTI and the other 3 receiving
DTI followed by mand training. Eye contact and challenging behaviors were also assessed across condi-
tions. Results indicate that 5 of 6 participants made more independent requests and acquired requesting
faster in the mand training condition, had slightly better eye contact in the DTI condition, and fewer
challenging behaviors in the mand training condition. Overall, the results indicate that mand training is a more
efficient method for teaching children with autism to make requests.
Key words: Mand training, motivating operations, discrete trial instruction, autism

One of the primary deficits in autism is a defi- Sundberg and his colleagues have advocated
cit in language and communication skills (APA, for mand training as an essential feature of early
2000). One particular difficulty for children with stages of a language training program for chil-
autism is making their needs and wants known, dren with autism (Sundberg & Partington, 1998)
or manding. This is a skill that develops early because the mand gives the child some control
in typically developing children, but children over the environment and increases the value
with autism often use other ways to get their of language as a form of social behavior
needs met, including engaging in problem be- (Sundberg & Michael, 2001). In addition, mands
haviors. Traditionally, behavior analytic meth- may be more likely to be used spontaneously
ods for teaching language and communicative because of the motivational factor and are there-
skills to children with autism have relied on dis- fore, in theory, more likely to occur under a va-
crete trial instruction (DTI; Koegel, Russo, & riety of environmental conditions (Sundberg &
Rincover, 1977). However, in recent applied Michael, 2001). Aside from these functional rea-
behavior analytic literature, there has been an sons, an analysis of the language of infants
increased focus on Skinners verbal operants suggests that mands are the first type of lan-
(Skinner, 1957) for teaching language to chil- guage to develop (Drash & Tudor, 1993).
dren with autism. For example, a set of teach- Despite the recent popularity of this strat-
ing procedures which relies on the arrangement egy and its supposed clinical utility, there is
of motivating operations (MOs) has been de- little research regarding this approach to teach-
veloped by researchers such as Sundberg and ing language in autism, particularly in compari-
Partington (1998). This procedure, known as son with more traditional teaching procedures
mand training, focuses on altering the MO in such as DTI. In one study, Drash, High, and
order to evoke verbal behavior that is under its Tudor (1999) investigated shaping a mand rep-
control. Laraway, Snycerski, Michael, and Pol- ertoire as the first step in language training in
ing (2003) have described MOs as stimuli that three nonverbal participants. A motivating op-
alter the value of an object or event as a rein- eration was contrived, the item was held out of
forcer (or punisher) and simultaneously alter reach, and the participant was asked if he
the frequency of a behavior that has been fol- wanted the item. At first any vocalization was
lowed by that reinforcer. reinforced, this was followed by differential re-
inforcement for certain imitated sounds, and
ultimately successive approximations to spe-
Correspondence information: Heather Jennett,
Kennedy Krieger Institute, Department of Behavior cific responses were reinforced. Since these
Psychology, 707 N. Broadway, Baltimore, MD items were visible and the sounds or words
21205; e-mail: were prompted, they were under the multiple

control of a nonverbal stimulus (as tacts) and that the NLP results in increased language
the experimenters verbal behavior (as echoics). (Koegel, Koegel, & Surratt, 1992; Koegel et al.,
Thus, beginning mand training was also echoic 1987), increased spontaneous language (Koegel
and tact training. Overall, the three partici- et al. 1987), and increased speech intelligibility
pants acquired mand and echoic repertoires (Koegel, Camarata, Koegel, Ben-Tall, & Smith,
and two of the participants began to acquire 1998). Furthermore, these comparisons yielded
a tact repertoire. Additionally, as manding results showing increased generalization
increased, inappropriate behavior and across people and settings (Koegel et al., 1987)
nonresponding decreased. as well as reduced disruptive behavior during
The Drash et al. (1999) study represents a teaching (Koegel et al., 1992) when taught us-
good first step, but more research is needed. ing the NLP.
Based on their findings, the authors discuss Thus, many of previous studies comparing
the value of initiating a language training pro- naturalistic teaching procedures with DTI have
gram with mand training rather than the tradi- found some advantage for the naturalistic pro-
tional imitative model (such as DTI), however, cedure. However, despite the remarkable re-
they do not provide control data using an imi- sults, there is a paucity of research in this area.
tation model as the first procedure. Another Particularly lacking is research focusing on the
drawback of the study was that the mands were application of Skinners analysis of verbal be-
always prompted (e.g., participant is asked havior to teaching language to children with
What do you want?) and thus it is unclear autism. Although mand training may resemble
whether the stimulus control for the response the NLP and incidental teaching in practice, nei-
was a motivating operation or a discriminative ther teaching procedure is conceptualized
stimulus or both. Despite these limitations, this based on Skinners analysis. However, because
study provided initial evidence of the clinical mand training procedures are similar to these
significance of mand training. other naturalistic procedures, the results can
Further, a few studies have investigated as- be expected to be the similar. Despite this, there
pects of language acquisition under the differ- are no published studies comparing mand train-
ent conditions of DTI and more naturalistic ing with DTI to date.
teaching methods, such as incidental teaching The purpose of the present study is to com-
(e.g., McGee, Krantz, & McClannahan, 1985) pare DTI with mand training for teaching chil-
and the Natural Language Paradigm (NLP; dren with autism to request items. Proponents
Koegel, ODell, & Koegel, 1987). Like mand of using Skinners analysis of verbal behavior
training, both incidental teaching and NLP ar- suggest that mand training should be the first
range the environment based on the motiva- step in a language program (Sundberg &
tion of the child. Although there may be minor Partington, 1998). Although Drash and col-
procedural differences, the main difference be- leagues (1999) claimed that their results sup-
tween these latter procedures and mand train- ported the superiority of mand training for lan-
ing is the conceptual basis. Investigations com- guage development over the traditional imita-
paring DTI with incidental teaching have fo- tion model, they did not directly compare these
cused on language skills such as answering methods. The present study investigated this
yes/no questions (Neef, Walters, & Egel, 1984), suggestion by initiating a language training
preposition use (McGee et al., 1985), and use of program through mand training with half of the
color adjectives (Miranda-Linne & Melin, 1992). participants, followed by language training in a
The results of these studies suggest that there discrete trial model. The other half of the par-
may be little difference in acquisition or reten- ticipants began their language training program
tion between the two teaching procedures through DTI, which was followed by mand train-
(McGee et al., 1985) but that DTI may enhance ing. Therefore, this study not only compares
quicker initial acquisition (Miranda-Linne & the effects of the two different teaching proce-
Melin, 1992). Additionally, DTI has been found dures for each participant but also investigates
to be more time efficient (Miranda-Linne & the effects of initiating a language program with
Lenin, 1992) but incidental teaching promotes each of the teaching models.
better generalization of skills (McGee et al., 1985; Several contrasts between DTI and natural-
Miranda-Linne & Lenin, 1992; Neef et al., 1984). istic procedures have been described in the lit-
Research comparing DTI with the NLP suggests erature (e.g., Delprato, 2001). In the present

Table 1
Participant Characteristicsa


Overall SC domain
Maggie 43 > 19 25 17
Marcus 511 210 34 25
Oliver 42 27 44 36
Peter 30 28 111 18
Christian 58 >19 40 25
Ryan 37 110 23 17
reported in yrsmos AE = age equivalent; SC = socialcommunication

study, all of these contrasts remained except (PPVTIII; Dunn & Dunn, 1997) was adminis-
for the element of specific versus nonspecific tered to obtain a standardized measure of each
reinforcement. That is, in both teaching proce- participants language ability; and the Scales
dures, the participants received specific, func- of Independent BehaviorRevised (SIBR;
tional rewards for a correct response and the Bruininks, Woodcock, Weatherman, & Hill,
items used in each procedure were items that 1996), an adaptive behavior scale with a pri-
were highly preferred by each individual par- mary focus on the participants social and com-
ticipant. Therefore, this study investigated the municative behaviors, was administered to the
role of the motivating operation rather than the parent of each participant as a checklist. Table
role of reinforcement. To do this, it was impor- 1 describes participant characteristics includ-
tant to keep the opportunities for reinforcement ing chronological age and age equivalent
constant across teaching procedures, but it was scores for the PPVTIII and SIBR. The par-
expected that the value of the item as a rein- ticipants were matched as well as possible by
forcer would differ depending on whether the their age equivalent score on the PPVTIII and
procedure follows the motivating operation (as each member of the matched pair was assigned
in mand training) or whether an item that has to one of the two conditions (DTI first or mand
been found to be preferred functions as a rein- training first). Assignment to a condition was
forcer for its request when it is chosen by the random but also depended upon which partici-
instructor (as in DTI). pant was ready to schedule training first.


PARTICIPANTS The stimulus materials consisted of toys and

activities that were made up of two parts and
Six participants with previous diagnoses of that needed both parts to be functional, such
Autistic Disorder and Pervasive Developmen- as paper and crayons. A standard pool of 24
tal Disorder, Not Otherwise Specified (PDD- items was established for the purpose of the
NOS) were recruited from a waiting list of chil- study and all items selected were popular toys
dren to receive services from the Douglass and activities for young children. From this
Developmental Disabilities Center, a specialized pool, two sets of materials were created: Set A
program for autism. Each child selected for par- consisted of one member of each pair and Set B
ticipation was requesting at a rate of less than was the other member. For the discrete trial
one per minute during a 20-min screening ses- condition, the participant was taught to request
sion. The six participants included five boys items from Set A and for the mand training con-
and one girl between the ages of 3 and 6. dition, to request items from Set B. For example,
During the intake procedure, the Peabody if the pair of objects was paper (A) and crayons
Picture Vocabulary Test Third Edition (B), requesting the paper was taught during


Sample List of Items
The effects of both DTI and mand training
Set A Set B were investigated through two concurrent mul-
Cassette tape Tape player tiple probe designs (Horner & Baer, 1978)
Mitt Softball across participants. In order to control for or-
der effects, half of the participants (Maggie,
View-Master Pictures Marcus, Oliver) were trained through mand
Racecars Car launcher training first followed by DTI, and the other
Markers Paper half (Peter, Christian, and Ryan) were trained
through DTI followed by mand training.
Bubbles Blower
Lite Brite Pegs PROCEDURES
Computer Game
Juice Straw Each participant received 1:1 instruction
during daily sessions consisting of either
Play-Doh Fun Factory
mand training or DTI. All sessions were 20
min in length and no more than two sessions
DTI and requesting the crayons was taught were conducted on a single day; typically, 8-
during mand training. A preference assess- 10 sessions were conducted per week. All
ment, based on the procedures of Fisher and sessions were conducted by the first author
colleagues (1992), was conducted for each par- in a small clinic room containing a table, chairs,
ticipant to determine his or her top 10 prefer- and adequate floor space. Procedures are de-
ences from the standard pool. Items were se- scribed in detail below and differences between
lected for each participant if they were chosen the procedures are summarized in Table 3.
at least 60% of the time relative to the other
items being assessed. If 10 pairs of items were Mand Training
not obtained from the standard pool of items,
an additional set of items was created con- During mand training, all of the stimulus ma-
taining items specific to the interest of that terials from Set A (nontarget items) were
participant, as reported by the participants placed around the room and were all within
parent or observed by the primary investiga- reach of the participant, who was permitted to
tor. See Table 2 for a sample list of items used. freely walk around throughout the session.
After each participants items were chosen, The participant initiated a trial by indicating
an assessment was conducted to determine interest in an object by approaching it, reach-
whether the participant could tact the items. If ing for it, pointing or using language to re-
the participant could tact items in his or her quest it. All of the Set B items were kept in an
set, a different label was given to the item. For opaque container, which was in view but out
example, for a participant who was able to tact of the reach of the participant. Correct re-
Thomas the train as Thomas, the item was sponses were reinforced with 30 s access to
labeled as train. For participants who were the pair of items. At the conclusion of the
observed to use more complex language, a reinforcement period, each member of the pair
more complex label was given to the item (e.g., was replaced to its original position so that
Thomas became Thomas the train). This the participant could initiate a subsequent trial.
was done to ensure that all participants were Trials were continued until 20 min had expired.
starting at comparable levels with their vocabu- Baseline. Once the participant indicated in-
lary and to reduce the chance that a participant terest in an item in the room, the instructor
who was able to tact more items prior to train- presented the target item (Set B) from the pair
ing would acquire the skill faster than a partici- and briefly played with them together before
pant who was unable to tact any items. Labels giving the initial item back to the participant.
for each member of the pair were of equal diffi- The instructor held back the second part with-
culty, in terms of number of syllables, so that out prompting for up to 5 s. If the participant
the language requirements for each participant made a correct response, he or she was given
in the two conditions would be comparable. access to the item for up to 30 s. If the partici-

Table 3
Procedural Differences Between DTI and Mand Training
DTI Mand training

Instructional Materials Selected by Instructor Selected by Child

Environmental Arrangements Child sits at table across from Child moves around
instructor the room with instructor

Controlling Variable Discriminative Stimulus Motivating Operation

(What do you want?)

Prompt Fading Progressive time delay Progressive time delay

across sessions within sessions

Mastery Criteria 80% correct across 1 mand per minute across

2 consecutive sessions 2 consecutive sessions

pant did not correctly respond after 5 s but removed from the array for the remainder of
was still indicating interest in the item (e.g., the training. This was to prevent the partici-
pointing, grabbing, otherwise vocalizing), he pant from focusing exclusively on one item
or she was also given access to the item but and to increase the variety of items for which
this was not considered a correct response. the participant could potentially request. The
This was done to approximate how parents removed item was replaced by the next high-
may have been reinforcing their childs mands est preferred item according to the
in the natural environment and also to pair the participants preference assessment and was
instructor with reinforcement from the begin- used as a replacement item for the remainder
ning of the study, as is recommended by of training. This occurred only one time (for
Sundberg & Partington (1998). Marcus) throughout training.
Training. The mand training procedures If a participant was playing with the nontar-
were set up and initiated exactly like the get item only and was not indicating interest
baseline assessment, however, the instructor in its partner target item, the nontarget item
held onto the target item and modeled the re- was removed from the participant and put back
sponse (e.g., I want crayon.). Successive into the array of items after 15 s. The instruc-
approximations of the response were followed tor attempted to contrive interest in another
by social praise and access to the item for 30 set of items. If a participant returned to the
s. If the participant made an error or did not same item and was playing inappropriately or
respond after 5 s but was still indicating inter- engaging in self-stimulatory behavior with the
est in the item, a second prompt was provided. item for more than 2 min, the pair was removed
As long as the participant was indicating in- from the remainder of that daily session and
terest in the item, prompts were continued ap- returned to the array for the next daily ses-
proximately every 5 to 10 s. After the partici- sion. This situation occurred only once (for
pant responded correctly to the instructors Maggie) throughout training. Finally, if the par-
model for a specific item at least 2 times within ticipant was not indicating interest in any of
a session, the prompt was faded during sub- the items, the instructor played with the pair to
sequent trials with that item. If the participant contrive interest and continued playing with
did not respond independently and was still different pairs of items until the participant in-
indicating interest in the item, a model was dicated interest. Requesting, as taught by mand
provided after approximately 5 s. training, was considered mastered when the
If a participant requested only one item participant was requesting at a rate of at least
across two consecutive sessions, this item was one per minute over two consecutive sessions.

Discrete Trial Instruction In contrast to the baseline procedures, initial tri-

als were immediately prompted with a zero time-
During DTI the participant was seated at a delay prompt (e.g., say, I want the ____. ).
table with the instructor. The instructor deter- Once the participant reached the criterion of 80%
mined when to present each stimulus item ac- correct over two consecutive sessions, the
cording to a randomly ordered list. Prior to each prompt was faded to a 2-s time delay, followed
training trial the instructor provided the partici- by a 5-s time delay after meeting the same crite-
pant with the item from the selected pair that he rion. Finally, after achieving a criterion of 80%
or she was not being taught to request (i.e., Set correct over two consecutive sessions with a 5-
B). One trial was conducted on each item before s time delay, the prompt was removed and only
moving onto the next item on the list. Once the used during the correction procedure. Correct
list was complete, the instructor repeated the responses were followed by social praise and
process from the beginning of the list. Trials were access to the requested item for 30 s. Errors or
conducted until 20 min had expired. Three differ- nonresponses were ignored during that trial and
ent lists of the random order of item presenta- a correction trial was implemented during which
tion were created for each participant and the the response was immediately prompted. Sub-
lists were rotated for each session to increase sequently, the missed trial was re-presented with
the probability that the items would be taught at the appropriate prompt level. After two incorrect
equal rates. trials, the next pair of items was presented and
Each trial was also interspersed with high- the item was re-introduced as designated by the
probability tasks at a ratio of 1:2. High-prob- list. Requesting, as learned through DTI, was
ability tasks varied among participants and considered mastered if the participant attained
such information for individual participants was 80% correct independent responding over two
obtained from their parents or through a brief consecutive sessions across ten pairs of items.
informal assessment. Because there may or may
not have been a motivating operation in effect Response Requirements
for the item selected by the instructor for the
training trial, there was a possibility of a lower Response requirements were comparable
level of reinforcement during DTI. Thus, the across conditions. Specific requirements were
purpose of interspersing trials with high-prob- determined by the instructor based on an infor-
ability items was to ensure a rich overall level mal assessment of the participants verbal reper-
of reinforcement in the discrete-trial condition. toire during the initial sessions. Requirements
Compliance resulted in high quality social praise ranged from the name of the item only, to the
or attention. name of the item with a carrier phrase (e.g., want
Baseline. After being presented with the (item name) or I want (item name)). Maggie,
nontarget item from Set B, the participant was Marcus, and Ryan were taught to request using
presented with its partner from Set A and asked I want (item name). Initially, Oliver was also
What do you want? No prompts were given. taught to request using I want (item name) but
A correct response was followed by access to his response requirement was reduced to just
the pair of materials for up to 30 s. Nonresponses the item name after he demonstrated difficulty
were ignored and the next pair of items was pre- with acquisition. Peter was taught to request
sented. If, after 5 s, the participant was indicat- using want (item name) and Christian was
ing a desire for the target item (Set A), such as by taught to use only the name.
reaching, pointing, grabbing, or vocalizing, the
participant was given access to the pair of items Dependent Variables
but the trial was marked as incorrect. This was
done to correspond with the similar procedure Frequency data were collected on two depen-
in the mand training baseline. dent variables: independent requests and echoic
Training. The training procedures were set requests. Independent requests were defined
up like the procedures in the baseline phase such as correct responses or approximations of the
that the participant was initially given the non- target word or phrase that occurred without a
target item from the pair. This was followed by model from the instructor; echoic requests were
the presentation of the target Set A item from defined as correct responses or approximations
that pair along with the SD What do you want?. of the target word that occurred within 5 s of the

instructors model. Additionally, in order to sessions for challenging behaviors and yielded
compare variety in requests between the two an average of 92.1% and 90.8%, respectively.
teaching conditions, data on the number of
different items requested (independent or Procedural Integrity
echoic) during each training session were col-
lected. Data were taken live by the instruc- In addition, procedural reliability was ob-
tor, with a portion of the sessions also coded served to ensure that the procedures for DTI
by a second independent coder for inter- and mand training sessions were implemented
observer agreement. as planned. An independent observer com-
In addition, eye contact and challenging be- pleted a checklist of items specifying the exact
haviors were observed as measures of social procedures that should be followed and rated
engagement. Both were coded from videotapes whether each component was completed prop-
by independent observers. Eye contact was de- erly. This was done for 10% of randomly ob-
fined as the participant directing his/her gaze served sessions across participants and across
toward the instructors face. Observations of eye training conditions. The average procedural
contact were coded using a 15-s momentary time integrity was 97.1% for DTI sessions and 97.7%
sample. Due to technical difficulties or camera for mand training. The main procedural compo-
obstructions, not all sessions could be coded. nent that was not consistently implemented
In total, 92.2% of DTI sessions (Maggie, 81.3%; correctly was the time allowed for reinforce-
Marcus, 76.5%; Oliver, 100%; Peter, 100%; Chris- ment. Specifically, reinforcement periods were
tian, 97.1%; Ryan, 100%) and 76.9% of mand occasionally 3540 s rather than 30 s and typi-
training sessions (Maggie, 71.4%; Marcus cally depended on the participants behavior at
62.5%; Oliver 75%; Peter, 100%; Christian, 100%; the time.
Ryan 81.8%) were coded. Challenging behavior
was defined broadly to capture the variety of Social Validity
behaviors exhibited by the participants and in-
cluded crying, screaming, hitting, throwing ma- Following the completion of the training, each
terials, running away from the instructional area, participating childs parent was provided with
noncontextual vocalizations, and self-stimulatory four randomly obtained videotaped samples of
behavior (e.g., tensing the body). Observations the training as a measure of social validation.
of challenging behavior were coded using a 1- Samples included two examples of both train-
min partial interval procedure. As with eye con- ing procedures, one from the first half of train-
tact, not all sessions could be coded due to tech- ing and one from the last half. Due to the par-
nical difficulties. In total, 96.1% of DTI sessions ents strong interest in the training methods
(Maggie, 100%; Marcus, 82.4%; Oliver ,100%; used, the entire 20-min session was included
Peter, 100%; Christian, 97.1%; Ryan, 100%) and for each participant, and parents were asked to
90.4% of mand training sessions (Maggie, 100%; watch at least 5 min of each session before rat-
Marcus, 75%; Oliver, 83.3%; Peter, 100%; Chris- ing it. They rated each sample on six dimen-
tian, 100%; Ryan, 91%) were coded. sions using a scale of 1 (not at all) to 5 (very
much). The dimensions included engagement,
Inter-Observer Agreement appropriate eye contact, appropriate communi-
cation, enjoyment, amount of learning, and
Inter-observer agreement (IOA) was collected whether the parents would use the teaching
on all variables. The event recording method procedure again in the future.
(IOA = (smaller frequency/larger frequency) x
100) was used for the requesting variables, and RESULTS
agreement data was collected across all partici-
pants and all conditions. During acquisition, REQUESTING VARIABLES
IOA was assessed for 53.2% of sessions and
yielded an average agreement of 93.7%. The Mand Training Followed by Discrete Trial
interval-by-interval method was used to col- Instruction
lect agreement data for both eye contact and
challenging behaviors. Data was collected for As shown in Figure 1, two of the three partici-
37.3% sessions for eye contact and 33.1% of pants receiving mand training prior to DTI

Figure 1. Independent and echoic requests in mand training followed by DTI (BL (A,B) = baseline for set
A & B items; BL (A) = baseline for set A items).

(Maggie and Marcus) made more indepen- ticipant (Oliver) made few or no independent
dent requests and required fewer sessions requests in either condition.
to meet criterion in the mand training condi- Maggie and Marcus. Neither Maggie nor
tion than in the DTI condition. The third par- Marcus made any independent requests for Set

Figure 2. Independent and echoic requests in DTI followed by mand training (BL (B,A) = baseline for set
B & A items; BL (B) = baseline for set B items).

B items during the initial mand training baseline. the mand training phase, Maggie made an av-
However, Maggie made an average of 1.25 erage of 15.3 independent requests (range 0
(range 03) and Marcus made an average of 1.5 70) and 14.4 echoic requests (range 423), al-
(range 05) independent requests for Set A though her performance was variable. She re-
items during the initial DTI baseline. During quired 14 sessions to meet the mastery crite-

rion for mand training. Marcus made an aver- which Oliver made an average of 1.0 indepen-
age of 13.7 independent requests (range 434) dent request (range 02) for Set A items.
and 14.4 echoic requests (range 727) with a Twelve sessions of DTI were conducted and
significant increasing trend as compared to the included the additional ASL prompts. How-
mand training baseline. He required 8 sessions ever, Oliver did not make any independent re-
of mand training to reach the mastery criterion. quests during any of these sessions. All
In the subsequent DTI baseline, Maggie twelve sessions were conducted at the 0-s time
made an average of 0.3 independent requests delay prompt level and his average percent-
(range 01) and Marcus made an average of 4 age correct was 48.5% (range 20.868.4%). His
independent requests (range 26) for Set A percentage correct at this prompt level ex-
items. In the DTI condition, Maggie made an hibited some variability and little increase.
average of 5.4 independent requests (range 0 This condition was also discontinued after
15) and 17.9 echoic requests (range 1524). She 12 sessions.
required 7 sessions at the 0-s time delay prompt
before reaching criteria for moving to the next DISCRETE TRIAL INSTRUCTION FOLLOWED BY
prompt level in the teaching procedure. Nine MAND TRAINING
additional sessions were conducted at the 2-s
time delay prompt level during which she As shown in Figure 2, all participants in
showed an initial increase, followed by a pla- this group (Peter, Christian, and Ryan) made
teau and then eventually a decrease. Maggie more independent requests and required
did not meet the criterion for mastery in the DTI fewer sessions to meet criterion in the mand
condition and made an average of 33.0% inde- training condition than in DTI. During the
pendent requests (range 20.845.5%) during the initial baseline conditions, all 3 participants
2-s time delay. Marcus made an average of 7.1 made 0 or near-0 independent requests. In
independent requests (range 012) and 16.9 the subsequent DTI condition, Peter made an
echoic requests (range 1222) for Set A items. average of 11 independent requests (range 0
He required the minimum of two sessions at the 20) and 9.7 echoic requests (range 2 20)
0-s time delay prompt level before reaching the across 6 sessions, by the end of which he had
criterion for moving to the next prompt level. met criteria for mastery in this condition.
Fifteen additional sessions were conducted at Christian made an average of 13.0 indepen-
a 2-s time delay during which his performance dent requests (range 021) and an average of
stabilized. Like Maggie, Marcus did not reach 17.3 echoic requests (range 825) during the
the mastery criterion for DTI before his perfor- DTI condition. Two sessions were conducted
mance stabilized. He made an average of 33.1% at the 0-s time delay prompt level followed by
of independent requests (range 2048%) at this 33 sessions at the 2-s time delay prompt level.
prompt level. Although Christians independent requests
Oliver. During the initial baseline condi- showed a slow steady increase, this condi-
tions Oliver made no independent requests tion was discontinued at the 2-s time delay
for the Set B items in the mand training baseline prompt level after his performance decreased
and an average of 0.25 independent requests slightly and he was no longer making
(range 01) for the set A items in the DTI progress. Ryan made an average of 8.7 inde-
baseline. During the mand training condition, pendent requests (range 0 21) and 15.2
Oliver made an average of 0.7 independent re- echoic requests (range 426). Eight sessions
quests (range 04) and 14.2 echoic requests were conducted at the 0-s time delay prompt
(range 923) for Set B items. Due to limited level followed by seven additional sessions
progress after 6 sessions of mand training, ad- at the 2-s time delay prompt level. This con-
ditional prompts were added in the form of dition was discontinued after Ryans perfor-
American Sign Language (ASL) and the re- mance stabilized and therefore he did not reach
sponse requirement was modified such that a the mastery criteria.
reduced vocal or sign response would be ac- In the subsequent mand training baseline,
cepted. Despite this, Oliver showed a negli- none of the participants made any indepen-
gible change compared to baseline and this dent requests for Set B items. During the mand
condition was discontinued. Three additional training condition, Peter made an average of
DTI baseline sessions were conducted during 17.5 independent requests (range 131) and an

Figure 3. Average number of items taught for each participant in DTI and mand training.

Figure 4. Percent of intervals with eye contact during DTI and mand training.

average of 10 echoic requests (range 5 16). 226) and an average of 20.6 echoic requests
He required 4 sessions to meet the mastery (range 1525) for Set B items and 11 sessions
criterion. For Christian, 3 sessions of mand were required before he met mastery criterion.
training were conducted during which he made
an average of 22 independent requests (range Number of Items
533) and an average of 26.3 echoic requests
(range 2032) for Set B items. Ryan made an All 6 participants requested more items in
average of 10.5 independent requests (range the DTI condition than in the mand training

Figure 5. Percent of intervals with challenging behaviors during DTI and mand training.

Figure 6. Social validity ratings from parents of participating children for DTI and mand training.

condition. On average, a participant learned average number and range of items learned
to request 9.6 items per DTI session and 3.4 per session for each participant.
items per mand training session. In DTI, all
the participants had the opportunity to re- EYE CONTACT AND CHALLENGING BEHAVIOR
quest all 10 items in at least half of the ses-
sions. In contrast, the maximum number of As shown in Figure 4, all participants had
items a participant learned to request in a mand more eye contact during DTI sessions than
training session was 6. Figure 3 shows the during mand training sessions. More specifi-

cally, the average percent of intervals with eye sessions. In contrast, the parents average
contact during DTI ranged from 6.7 % to 18.5 ratings of eye contact favored DTI (x=3.38)
%, whereas the average percent of intervals over mand training (x=3.25).
with eye contact during mand training ranged
from 0.5 % of intervals to 14.5 %. As shown in DISCUSSION
Figure 5, two of the participants (Marcus and
Oliver) engaged in more challenging behavior In this study, the effects of DTI and mand
during DTI than during mand training while training on the acquisition of requests were
the other four engaged in approximately equal compared using concurrent multiple probe de-
levels across conditions. Specifically, Marcus signs across participants. Each training proce-
engaged in challenging behavior during an av- dure was implemented as it is typically con-
erage of 16.3% of intervals (range 535) dur- ducted while controlling for opportunities for
ing DTI, and an average of 0.8% of intervals reinforcement and preference in order to inves-
(range 05) during mand training. The primary tigate the role of the motivating operation. The
topographies of his challenging behaviors results suggest that, although both mand train-
were running away from the instructional area ing and DTI facilitate acquisition of requests in
and throwing materials. Oliver engaged in chal- participants with autism, mand training facili-
lenging behaviors during an average of 46.6% tates increased spontaneous requesting and at
of intervals (range 1090) during DTI, and a faster pace. Five out of the six participants
during an average of 35% of intervals (range made more independent requests for items when
1050) during mand training. During DTI and the motivating operation was considered than
mand training, the primary topographies of when preferred items were delivered contingent
Oliver s challenging behaviors were upon correct requests for which there may or
noncontextual vocalizations and crying. He may not have been a motivating operation in
also engaged in tensing during DTI but not effect. In addition to making more requests,
during mand training. The other four partici- these five participants also met the mastery cri-
pants rarely exhibited challenging behaviors teria in fewer sessions in the mand training con-
but when they did the primary topographies dition than in the DTI condition.
were running away from the instructional area, Although five participants showed the pat-
throwing materials, and crying, and these oc- tern described above, perhaps it is best high-
curred equally across conditions. lighted by Christians performance. In the dis-
crete trial condition, Christian acquired the skill
Social Validity at a slow and steady pace and after 35 ses-
sions, this condition was discontinued due to
The videotaped samples and social validity the stability of his responding. In contrast, he
rating scales for each participant were sent to met mastery criteria for the mand training con-
all of the participating parents except Olivers. dition in only three sessions and made an aver-
A clinical decision was made not to send the age of nine more independent requests per ses-
samples to his parents since he did not make sion than in the DTI condition. Prior to train-
any progress. All of the other parents com- ing, Christian did not engage in any vocaliza-
pleted the surveys. Data are reported in Fig- tions related to the items; he typically pointed
ure 6 and show that parents rated both teach- to or grabbed them. It is possible that his rapid
ing methods favorably but tended to favor increase in responding in the mand training
mand training slightly over DTI on all dimen- condition was primed by his recent history of
sions except eye contact. Specifically, aver- reinforcement for vocal responses in the dis-
age ratings indicate that parents rated their crete trial condition. However, if reinforcement
childs engagement (DTI, x=4.25, mand train- alone were responsible for the increase in his
ing, x=4.5), childs enjoyment (DTI, x=4.13, responding, one would expect to see a similar
mand training, x=4.88), childs appropriate increase in the discrete-trial condition at some
communication (DTI, x=4.25, mand training, point in the training. This and the fact that a
x=4.5), childs learning (DTI, x=4.38, mand train- similar pattern of results was obtained for four
ing, x=4.88), and their own desire for future other participants suggests that the motivat-
use (DTI, x=4.63, mand training, x=5) slightly ing operation was the primary factor in his
higher in mand training sessions than in DTI speed of acquisition. Thus it is hypothesized

that Christian would show a similar pattern of learning opportunities were controlled by the
results if the order of training were reversed. instructor. In contrast, the mand training con-
Across both teaching conditions, preference dition represented a free operant condition
and opportunities for reinforcement were held such that the participant and the motivating
constant since the same highly preferred stimuli operation controlled the pace and the number
were used in both conditions and consequences of learning opportunities. Thus, Peter was pos-
for a correct response were specific to the re- sibly able to acquire the skill more quickly in
sponse. However, it can be assumed that the the mand training condition because he was in
mand training condition involved richer rein- control of the pace and the number of
forcement because focusing on the motivating learning opportunities.
operation, by definition, guaranteed that more In principle, a free operant setting allows for
desirable items were delivered in this condi- a greater number of learning opportunities be-
tion. In contrast, because there may or may not cause there is no ceiling on the number of re-
have been a motivating operation in effect for sponses made by the learner. While this gener-
the items in the DTI condition, it can be as- ally favors increased responding for many learn-
sumed that some items may have had less value ers, it may also allow for a decreased number of
during their trials and therefore may not always learning opportunities for a child whose natu-
have functioned as reinforcers. Anecdotal ob- rally occurring motivating operation maybe
servations from the DTI condition indicated more limited compared with what the pace of an
that the majority of the participants sometimes instructor may provide or for a learner with a
indicated interest for other items aside from the strong motivating operation for inappropriate
one that was selected by the instructor for that behavior, such as self-stimulation or
trial. A few participants cried or otherwise ex- tantrumming. In these situations, the learner
pressed anger and frustration when trials were may benefit from more structured teaching
conducted with items in which they were not paced by an instructor or from more creativity
interested at the time or when items for which by the instructor in contriving other motivat-
there was a strong motivation were removed ing operations or shifting the childs focus.
and a new item was presented. In the mand Although five of the six participants acquired
training condition, the participant was able to different levels of requesting in both mand train-
continue playing with the item for as long as ing and DTI, Oliver did not acquire the skill in
there was a motivating operation. either condition. Initially, the response required
Another factor likely to facilitate increased of him was I want (item name). As with all the
responding in a mand training session is that participants his response requirement was de-
following the motivating operation allows for termined based on an informal assessment of
an increased number of learning opportunities. his language abilities during the initial assess-
This is best demonstrated by Peters perfor- ment period. During this time he was observed
mance. Although he showed the pattern of to speak in several-word phrases and some full
making more requests and requiring fewer ses- sentences, thus, he was erroneously assessed
sions to meet criterion in the mand training to have the ability to produce a full sentence to
condition he was the only participant to meet make requests. However, after six sessions of
criteria in the DTI condition. He met the prompt- mand training, Oliver was not making any inde-
fading criteria in the minimal amount of ses- pendent requests. It is notable that he also
sions and was independently requesting at the tended to require about three or four modeled
desired level by the time he was receiving the responses in order to emit an echoic request.
5-s time delay prompt. The slope of his acqui- Sometimes he persisted through these multiple
sition curves across the two conditions is about prompts and at other times his focus shifted
equal. Peter did as well as could be done in the elsewhere. It was hypothesized that imitation
discrete trial condition but still made more re- was difficult for him and an informal imitation
quests and met the mastery criteria faster in assessment indicated poor performance with
the mand training condition. His performance general imitation skills. In a 10-min assessment
across conditions highlights the difference be- of gross motor and vocal imitation with dense
tween a restricted operant and a free operant. reinforcement, Oliver was only able to imitate
The discrete trial condition was a restricted 47% of the tasks. As a result, the response re-
operant such that the pace and the number of quirement was decreased to the name of the

item and additional prompts were added so that mented, which suggests that this language in-
Oliver would have a greater opportunity of con- tervention procedure is a more efficient method
tacting reinforcement more quickly. In addition for teaching children to request items. Although
to the typical verbal models provided for all the the results also indicate that acquisition using
participants, an American Sign Language (ASL) DTI is slower than with mand training, the de-
sign representing the desired items was used sign of the current study does not allow us to
following a most-to-least prompting strategy. determine whether performance in one condi-
Thus, Oliver was first physically prompted to tion has an impact on the speed of the subse-
emit the sign while the vocal response was quent condition. Future studies using an alter-
modeled simultaneously, then the sign and the nating treatments design might provide more
vocal response were modeled simultaneously, information regarding the speed of acquisition
followed by no prompts. Despite this, indepen- by directly comparing the two conditions and
dent requesting did not increase and remained might answer the question as to efficiency.
at near zero levels. Finally, this study investigated whether there
Although no formal data were taken on en- would be more eye contact and fewer challeng-
gagement with the materials, anecdotally Oliver ing behaviors during mand training than dur-
was observed to be almost continuously en- ing DTI. With regard to eye contact, the results
gaged with the target stimuli during the baseline are counter to the hypothesis that there would
period. He also indicated a desire (e.g., by grab- be more eye contact during mand training than
bing) to regain access to the materials once during DTI. All of the participants in this study
they were removed. However, once training had had previous experience with DTI and the
begun, his engagement with the materials and preacademic skills, such as sitting in a chair
nonvocal mands to regain access to them de- and attending to materials that are typically
creased. His engagement was replaced by chal- taught with DTI. It is possible that this history
lenging behaviors, particularly noncontextual and the common practice in DTI of establish-
vocalizations and tensing his upper body. Data ing the childs attention prior to issuing an in-
collected on challenging behaviors indicate that struction inflated the amount of measured eye
prior to any training, Oliver engaged in chal- contact during DTI. It is also notable that dur-
lenging behaviors during an average of 12% of ing DTI, the participant was usually sitting
intervals during the DTI baseline and during across from the instructor during teaching
an average of 16% of intervals during the mand whereas in mand training, the participant could
training baseline. However, as noted in the re- be anywhere in relation to the instructor. There-
sults section, this level increased to about 35% fore, there may be more opportunities for eye
during mand training and 47% during DTI and contact in DTI based on its structure. The hy-
the trend indicated a steady increase through- pothesis that participants would engage in more
out training. This increase may have occurred eye contact during mand training was based
naturally, but another possibility is that the in- on the belief that the instructor would be paired
troduction of the demand to imitate language, more with reinforcement during the mand train-
which was difficult for him, altered the motivat- ing condition and therefore would be more so-
ing operation away from engaging with the tar- cially desirable. However, many of the parents
get stimuli and toward engaging in other rein- who rated the social validity of these sessions,
forcing activities (e.g., self-talk, tensing). rated mand training slightly lower than DTI for
In addition to the differential impact of each appropriate eye contact and many noted that
teaching method on the acquisition of inde- their children appeared to be looking at the toys
pendent requests, this study sought to explore more than at the instructor.
the suggestion that children with autism should With regard to challenging behaviors, the
receive mand training as the initial intervention results provide some support for the hypoth-
(Drash et al, 1999; Shafer, 1994; Sundberg & esis that the participants would engage in fewer
Partington, 1998). This was accomplished by challenging behaviors during mand training.
training half of the participants with mand train- Two of the participants fit this pattern while the
ing followed by DTI and the other half in the other four engaged in approximately equal lev-
reverse order. The results indicate that mand els of challenging behavior irrespective of the
training provides faster acquisition of the skill teaching condition that was in place. None of
regardless of the order in which it is imple- these six participants had major behavior prob-

lems and perhaps the results with children who items more consistently due to increased op-
have more severe problem behavior would be portunities to practice but that in DTI they
more useful. However, both Marcus and Oliver learned to request a greater number of items
engaged in challenging behaviors during DTI but inconsistently. Future research should
that interfered significantly with their learning compare the two types of teaching procedures
because these behaviors were incompatible while controlling for the number of items that
with the teaching procedures. are being taught.
During the mand training condition, most of This study focused on teaching children with
the participants exhibited challenging behav- autism to request preferred items using mul-
iors such as leaving the instructional area, pos- tiple prompts and thus represents one of the
sibly because of a shift in the motivating op- first steps in mand training as outlined by
eration. Future studies or applied clinical prac- Sundberg and Partington (1998). That is, the
tice could modify the procedures of this study mands being taught were not pure mands which
to expand the range of items and activities that are controlled only by an establishing or moti-
children with autism are being taught to mand. vating operation. Rather, they were multiply
This could possibly reduce the level of chal- controlled by the motivating operation, tact
lenging behavior during mand training found prompts (i.e., the presence of the item) and of-
in the current study even further. However, simi- ten by echoic prompts (i.e., the modeled re-
lar modifications are more difficult to make in a sponse). The requests being taught in the DTI
DTI session because the additional behavior is condition were controlled by the discrimina-
incompatible with the teaching procedure where tive stimulus (i.e., the instruction What do you
the instructor is making the decisions of what want?) but also by tact and echoic prompts.
items to present and teach. Future studies could also investigate the ac-
This study also demonstrated clinical sig- quisition of pure mands such that the item be-
nificance, as shown through the social validity ing taught is not in the view of the learner but
ratings. In addition, at various points in train- rather hidden so that the true influence of the
ing, four of the six participants parents com- motivating operation can be investigated.
mented that their children were spontaneously
requesting more for items or activities at home, REFERENCES
something that was not observed prior to the
training sessions. American Psychiatric Association (2000).
Based on the results of this study, advan- Diagnostic and Statistical Manual of
tages of following the motivating operation in- Mental Disorders (Text Revision). Wash-
clude faster acquisition, a greater number of ington DC: Author.
requests per time period, and fewer challeng- Bruininks, R.H., Woodcock, R.W., Weather-
ing behaviors. The advantage of the teacher- man, R.F., & Hill, B.K. (1996). Scales of
directed and paced instruction, as set up in the Independent BehaviorRevised. Itasca,
study, is that the learner is being exposed to IL: The Riverside Publishing Company.
more stimuli and could potentially learn to re- Delprato, D. J. (2001). Comparisons of dis-
quest a greater variety of items. In the present crete-trial and normalized behavioral lan-
study the teaching procedures were imple- guage intervention for young children
mented as they are typically and the results with autism. Journal of Autism and De-
indicated that in a free operant situation chil- velopmental Disorders, 31, 315325.
dren with autism may focus on only a few items Drash, P.W., High, R.L., & Tudor, R.M. (1999).
at a time. Therefore, it is possible that the quicker Using mand training to establish an echoic
acquisition during the mand training procedure repertoire in young children with autism.
was because the participants were not required Analysis of Verbal Behavior, 16, 2944.
to learn to request every item in the set. It may Drash, P.W., & Tudor, R.M. (1993). A func-
have been easier for the participants to meet tional analysis of verbal delay in pre-
the criteria in the mand training condition be- school children: Implications for preven-
cause it did not specify that a certain number of tion and total recovery. Analysis of Ver-
items needed to be requested correctly, as in bal Behavior, 11, 19 29.
the DTI condition. It is possible that in mand Dunn, L.M., & Dunn, L.M. (1997). Peabody
training the participants learned to request fewer Picture Vocabulary TestThird Edition.

Circle Pines, MN: American Guidance Ser- terms to describe them: Some further refine-
vice, Inc. ments. Journal of Applied Behavior Analy-
Fisher, W., Piazza, C. C., Bowman, L. G., sis, 36, 407414.
Hagopian, L. P., Owens, J. C., & Slevin, I. McGee, G.G., Krantz, P.J., & McClannahan, L.E.
(1992). A comparison of two approaches for (1985). The facilitative effects of incidental
identifying reinforcers for persons with se- teaching on preposition use by autistic chil-
vere and profound disabilities. Journal of dren. Journal of Applied Behavior Analy-
Applied Behavior Analysis, 25, 491498. sis, 18, 1731.
Horner, R.D., & Baer, D.M. (1978). Multiple- Miranda-Linne, F., & Melin, L. (1992). Acquisi-
probe technique: A variation of the multiple tion, generalization, and spontaneous use
baseline. Journal of Applied Behavior of color adjectives: A comparison of inci-
Analysis, 11, 189196. dental teaching and traditional discrete-trial
Koegel, R.L., Camarata, S., Koegel, L.K., Ben- procedures for children with autism. Re-
Tall, A., & Smith, A.E. (1998). Increasing search in Developmental Disabilities, 13,
speech intelligibility in children with autism. 191210.
Journal of Autism and Developmental Dis- Neef, N.A., Walters, J., & Egel, A.L. (1984). Es-
orders, 28, 241251. tablishing generative yes/no responses in
Koegel, R.L., ODell, M.C., & Koegel, L.K. developmentally disabled children. Jour-
(1987). A natural language teaching para- nal of Applied Behavior Analysis, 17, 453
digm for nonverbal autistic children. Jour- 460.
nal of Autism and Developmental Disorders, Shafer, E. (1994). A review of interventions to
17, 187200. teach a mand repertoire. Analysis of Verbal
Koegel, R.L., Koegel, L.K., & Surratt, A. (1992). Behavior, 12, 5366.
Language intervention and disruptive be- Skinner, B.F. (1957). Verbal Behavior.
havior in preschool children with autism. Englewood Cliffs, NJ: Prentice Hall, Inc.
Journal of Autism and Developmental Dis- Sundberg, M.L., & Michael, J. (2001). The ben-
orders, 22, 141153. efits of Skinners analysis of verbal behav-
Koegel, R.L., Russo, D.C., & Rincover, A. (1977). ior for children with autism. Behavior Modi-
Assessing and training teachers in the gen- fication, 25(5), 698724.
eralized use of behavior modification with Sundberg, M.L., & Partington, J.W. (1998).
autistic children. Journal of Applied Be- Teaching Language to Children with Au-
havior Analysis, 10, 197205. tism or Other Developmental Disabilities.
Laraway, S., Snycerski, S., Michael, J., & Pol- Pleasant Hill, CA: Behavior Analysts, Inc.
ing, A. (2003). Motivating operations and