- Email: [email protected]
Can the attention training technique turn one marshmallow into two? Improving children's ability to delay gratification
Accepted Manuscript Can the Attention Training Technique Turn One Marshmallow into Two? Improving Children’s Ability to Delay Gratification Joanne Murray, Anna Theakston, Adrian Wells PII:
S0005-7967(15)30058-9
DOI:
10.1016/j.brat.2015.11.009
Reference:
BRT 2936
To appear in:
Behaviour Research and Therapy
Received Date: 25 June 2015 Revised Date:
19 November 2015
Accepted Date: 19 November 2015
Please cite this article as: Murray, J., Theakston, A., Wells, A., Can the Attention Training Technique Turn One Marshmallow into Two? Improving Children’s Ability to Delay Gratification, Behaviour Research and Therapy (2015), doi: 10.1016/j.brat.2015.11.009. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
RI PT
Can the Attention Training Technique Turn One Marshmallow into Two? Improving
Joanne Murray 1
M AN U
Anna Theakston 2
SC
Children’s Ability to Delay Gratification
Adrian Wells 1
1
University of Manchester, School of Psychological Sciences, Rawnsley Building, MRI
University of Manchester, School of Psychological Sciences, Coupland 1 Building, Oxford
EP
Road, Manchester M13 9PL, UK
AC C
2
TE D
Manchester, M13 9WL UK
Corresponding author: Adrian Wells1 Email: [email protected] Tel: 0161 276 5399
1
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT Abstract
The seminal Marshmallow Test (Mischel & Ebbesen, 1970) has reliably demonstrated that children who can delay gratification are more likely to be emotionally stable and successful
RI PT
later in life. However, this is not good news for those children who can’t delay. Therefore, this study aimed to explore whether a metacognitive therapy technique, Attention Training (ATT: Wells, 1990) can improve young children’s ability to delay gratification. One hundred children participated. Classes of 5-6 year olds were randomly allocated to either the ATT or a
SC
no-intervention condition and were tested pre and post-intervention on ability to delay
M AN U
gratification, verbal inhibition (executive control), and measures of mood. The ATT intervention significantly increased (2.64 times) delay of gratification compared to the nointervention condition. After controlling for age and months in school, the ATT intervention and verbal inhibition task performance were significant independent predictors of delay of gratification. These results provide evidence that ATT can improve children’s self-regulatory
TE D
abilities with the implication that this might reduce psychological vulnerability later in life. The findings highlight the potential contribution that the Self-Regulatory Executive Function
processes.
EP
(S-REF) model could make to designing techniques to enhance children’s self-regulatory
AC C
Keywords: Children, Attention Training, Metacognition, Self-regulation, Delay of Gratification.
2
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
Can the Attention Training Technique turn one marshmallow into two? Improving children’s ability to delay gratification In order to delay gratification, people need to be able to employ self-regulation. That is,
RI PT
the ability to override and change their response to an immediate impulse or desire. Studies have shown that children who are able to self-regulate in this way are likely to experience greater success later in life (Mischel, Shoda, & Peake, 1988; Shoda, Mischel, & Peake, 1990; Ayduk et al., 2000; Eigsti et al., 2006). Conversely, inability to delay gratification in
SC
childhood is a risk factor for subsequent psychological pathology (Moffitt et al., 2011). Given
M AN U
this consistently well-supported connection, it is worthwhile to explore whether ability to delay gratification can be improved during childhood with a goal of potentially enhancing outcomes later in life. However, research has tended to overlook this extension, focusing instead on the contexts in which children are best able to delay.
TE D
Delay of gratification is a dimension of self-regulation that can be seen as a correlate of executive control (Mischel, Yuichi, & Rodriguez, 1989). In a separate line of research, studies have aimed to improve components of executive function in children (Karbach &
EP
Kray, 2008; Minear & Shah, 2008). Rueda, Checa, & Combita (2012) delivered 10 sessions of computerised attention training to healthy five year old, Spanish children in school.
AC C
Children who had received the attention training displayed faster activation of brain areas associated with executive control. Further research with four to six year old German children (Streb, Hille, Schoch, & Sosic-Vasic, 2012) delivered computerised attention training every day for one week within school. Results indicated that children who had received the intervention displayed significant improvements in inhibition and cognitive flexibility. These studies suggest that attention training strategies in young children from non-clinical samples is feasible and may be advantageous. However, these studies stand as examples of a limited 3
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
number of studies conducted in this area; they have used small samples and not specifically measured self-regulation using the marshmallow task, which we know to be a predictor of longer term outcomes.
RI PT
The attention training delivered in previous research is time consuming for a classroom setting. For example, Rueda et al’s (2012) intervention lasted seven and a half hours over five weeks. Interventions also consisted of a wide range of components intended to target various aspects of executive functioning (e.g. cognitive flexibility, inhibition, tracking, discrimination
SC
or conflict resolution). Whilst this provides a comprehensive intervention, this is challenging
M AN U
for schools to implement within their existing curriculum. Of particular importance, such multicomponent techniques are not based on any specific model linking attentional control, delay of gratification and psychological vulnerability, which is our aim. We considered how a metacognitive therapy intervention, the Attention Training
TE D
Technique (ATT; Wells, 1990) could enhance children’s ability to delay gratification. This technique is based on a specific metacognitive model, the Self-Regulatory Executive Function (S-REF) model, linking executive control to unhelpful thinking styles that cause
EP
psychological vulnerability (Wells & Matthews, 1994; 1996). Specifically, Wells and Matthews argue that psychological dysfunction is caused by perseveration of unhelpful
AC C
processing that occurs as a cognitive attentional syndrome. For example, perseveration of ideational processes is evident in addictive behaviours where repetitive thinking is characterised not only by worry and rumination but also ‘desire thinking’ (Spada, Caselli, & Wells, 2013). This thinking style prolongs internal aversive experiences leading to greater behaviours aimed at escaping anxiety, low mood or desire. The S-REF model has led to the development of treatment techniques intended to increase control over extended thinking and increase flexibility in responding to internal states. The ATT aims to enhance the capacity to 4
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
disengage perseverative processing by enhancing executive control through training individuals in externally-focused auditory attention exercises (Wells, 2000). To date, the ATT has been delivered to adults, with studies indicating its effectiveness in improving symptoms in adult patients with mood disorder (Papageorgiou & Wells, 2000), social phobia (Wells,
RI PT
White, & Carter, 1997), and traumatic stress symptoms (Callinan, Johnson, & Wells, 2015; Nassif & Wells, 2014). Significant effects have been found after 2-6 sessions (Siegle, Ghinassi, & Thase, 2007; Papageorgiou & Wells, 2000).
SC
Whilst previous studies have focused on adults experiencing psychological disorder, it is
M AN U
plausible that the ATT could disrupt perseverative processing routines that contribute to low impulse control in children. Because the ATT is designed to disconnect sustained processing and coping efforts from internal and external events it is likely that children who are trained in it could subsequently develop enhanced awareness of the independence between internal and external experiences and behaviour and increase their flexible choice over action. An
TE D
important marker for this effect in children would be an enhanced ability to delay gratification (i.e. choose not to respond to perseverative, desire related thinking). If this effect
EP
can be demonstrated then an exciting implication is that ATT may be a candidate strategy for improving psychological wellbeing (i.e. resilience) throughout life.
AC C
Aims of the Present Study
Research indicates that children begin to develop inhibitory control between 5-6 years of age (Carlson & Moses, 2001; Nelson & Narens, 1994), and therefore the primary aim of this study was to explore whether it is possible to use ATT in young children aged 5-6 years, and whether this impacts on their subsequent ability to delay gratification. It was hypothesised that those children receiving ATT would be better able to delay gratification than those in a no-intervention control condition. The testing of this hypothesis was our primary objective. 5
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
A secondary aim was exploratory and examined whether children’s ability to delay gratification was correlated with performance on a verbal Stroop task (the day/night task: Gerstadt, Hong, & Diamond, 1994), prior to and following the ATT intervention, and
RI PT
whether, in line with previous research in adults, ATT improves mood in children. Performance on the day/night task has been shown to be dependent on the development of inhibitory control; we predicted positive associations between performance on the day/night task and delay of gratification. Furthermore, we intended to use scores on this task to control
SC
for individual differences in inhibitory control when testing the effects of ATT on delay of
M AN U
gratification. Finally, we wanted to explore any effect of the ATT on children’s performance on the day/night task following the intervention.
The following specific predictions and research questions were tested: •
Primary prediction: That 5-6 year old children who receive the ATT will be more
condition.
Secondary Questions: Are children who score higher on a verbal Stroop task more
EP
•
TE D
able to delay gratification at follow-up testing than children in the no-intervention
able to delay gratification? Will children in the ATT condition have significantly
AC C
improved mood and improved performance on the day/night task compared with those in the no-intervention condition? Materials and Method
Design Within this mixed-model design, classrooms of children were randomly allocated as blocks to either the experimental condition (where children received the ATT intervention, 6
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
described below) or the no-intervention condition (where children received school activities as normal). The same seven-day study protocol was used in each school. Day one (always a Monday) involved the researcher attending school to collect baseline data. Day two to five (always Tuesday to Friday) involved the teacher administering the intervention on three
RI PT
occasions, at their convenience (or no-intervention). There was then a two day weekend
break (Saturday and Sunday) followed by the researcher returning to school on day seven (Monday) to collect follow up data.
SC
The primary dependent variable was children’s ability to delay gratification and resist a
M AN U
treat (candy), measured by the Marshmallow Test (Mischel & Ebbesen, 1970). The control/predictor variables for the main analysis were child’s age (measured in months); number of months the child had been in school; child’s mood score measured at T1 and T2 by the Faces Scale (Holder & Coleman, 2008), and; child’s ability to suppress verbal
TE D
responses, measured by the day/night task at T1 and T2 (measures are described below). Recruitment. Five classes of children were recruited from five primary schools within Greater Manchester, England. Recruitment began in summer 2012 – spring 2013 via emails
EP
to school Head Teachers. Teachers then distributed information packs and consent forms to parents. For children to be eligible to participate, they had to have parental consent and be
AC C
able to read, write and understand English. The study was approved by the University of Manchester’s Ethics Committee (Ref – 12203). Participants. One hundred children from five classes participated in the study, 59 (three classes) were randomly allocated to the experimental condition, and 41 (two classes) to the no-intervention condition. Fifty eight children were male; 42 were female. Children’s age ranged from 5.20 years to 6.52 years (M = 5.87 years, SD = 0.3 years). The time children had been in primary school at baseline testing ranged from 2 to 7 months (M = 4.52 months, SD = 7
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
1.8). One participant was withdrawn from the study due to being absent from school at T2 testing. Randomisation. Within the experimental (ATT) condition, teachers were required to
RI PT
administer the intervention to their whole class therefore there was no option to randomly assign individual children within each class to each condition. Block randomisation was used in which a whole class was randomly assigned to either the experimental or no-intervention condition. A research randomiser tool (www.randomizer.org) was used to ensure that each
SC
class had an equal chance of being allocated to the experimental or no-intervention condition.
used to determine the allocation. Apparatus
M AN U
Once a class had confirmed their participation in the study, the research randomiser tool was
Attention Training Technique. A recorded version of Wells’ (1990) Attention Training
TE D
Technique was used and standardised instructions for delivery were provided to class teachers. They were instructed to play an ATT recording on any three separate occasions during the four day intervention period. The ATT consists of approximately 11 minutes of
EP
sounds (such as bird song, traffic, clock ticking) and a voice which instructs participants to
AC C
direct attention in specific ways, for example from one sound to another and to different spatial locations.
Materials. Schools provided a room for data collection to take place undisturbed. Chairs and tables were arranged to ensure the child would be positioned in front of the treat and so that the written activities could be undertaken. For the primary outcome of the Marshmallow Test, there were two types of treat, a chocolate button or a Haribo sweet (both vegetarian) to ensure that the treat was desired by 8
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
the child. A stopwatch was used to time the 13 minutes the child had to wait alone in the room. The time started when the door was closed. A standardised score sheet was used which also allowed for recording of verbal comments made by the child during data collection.
RI PT
The following measures were also administered; to assess executive control the ‘day/night’ task was used. This consists of a standardised set of 20 day and night cards (four practice cards and 16 test cards, as described by Gerstadt et al., 1994). The dimensions of each card were 13.5 x 10 cm. The participant is instructed to say “day” when a picture of the
SC
moon is presented and “night” when the sun is presented and the number of errors are
M AN U
measured. To assess mood the Faces Scale was used (Holder & Coleman, 2008). This comprises of a five-point self-report scale and children are asked to select the face which best represents their thoughts and feelings in response to three questions: 1) “most of the time I feel;” 2) “compared to other people I feel;” 3) “right now I feel”.
TE D
Procedure
All data was collected by the researcher or a volunteer Research Assistant who had been
EP
trained in the procedure.
Baseline testing session (day one). Children’s assent was gained using a child-friendly
AC C
information sheet. Children were then asked to complete the adapted Faces Scale; the researcher provided instructions and read out the three questions individually, pausing for the child to colour-in their response. The researcher then introduced the day/night task, explaining that the child should try to answer as quickly as possible. The child was instructed to say “day” when the moon card was presented and “night” when the sun card was presented. To ensure the child understood the rule, two practice trials were undertaken (see Gerstadt et al., 1994). 9
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
To ensure results were comparable with the original study, the testing phase of the day/night task involved the presentation of 16 cards (eight day and eight night) in the random order used in the original study (Gerstadt et al., 1994): night (n), day (d), d, n, d, n, n, d, d, n,
RI PT
d, n, n, d, n, d. The cards were presented at one second intervals. The researcher then presented the child with an option of a chocolate button or a Haribo candy and placed their selection on a plate directly in front of them. The researcher told the child: “you can eat this sweet now, or if you wait until I come back, you will get two sweets”.
SC
The researcher checked the child’s understanding and left the child alone in the room, using
M AN U
the stop-watch to time 13 minutes. All distractions had been removed from the room and they were given no further direction of what to do whilst they waited. The child was positioned with their back to the door so the researcher could see them through a window in the door. If the child ate the sweet or left the room, the researcher would thank them for participating and return them to their class. If 13 minutes elapsed and the child had not eaten the sweet, the
TE D
researcher would re-enter the room and give the child another treat. Intervention stage (day two to five). Children in the no-intervention condition continued
EP
with their usual school activities between baseline and follow-up.
AC C
Children in the experimental condition also continued with their usual class activities but also received the ATT as a whole class on three separate days, chosen at the convenience of their class teacher. The ATT was delivered by either the class teacher, or a teaching assistant, using standardised instructions. Follow-up testing (day seven). The researcher returned to school seven days later to undertake follow-up testing. The Faces Scale, day/night task and the Marshmallow test were administered using the same protocol as the previous week. During the follow-up testing, the 10
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
researcher made no direct reference to the ATT to reduce demand characteristics or potential for a placebo response from the children. Results
RI PT
Pre-Study Differences In the experimental condition there were 31 males and 28 females. In the no-intervention condition there were 27 males and 14 females. Table 1 shows the mean age, months in school
SC
and scores for each condition. At T1, children in the ATT condition had been in school longer (M=5.08) on average, than children in the no-intervention condition (M=3.71). Children in
M AN U
the ATT condition were on average, slightly younger (M=5.38 years) than children in the nointervention condition (M=5.92 years). Differences in age and months in school were due to the research being conducted throughout the 2012-2013 school year.
TE D
The Effect of ATT on Delay of Gratification
Chi-Square Tests were run to explore the difference in ability to delay gratification (DG) in the no-intervention and experimental condition at baseline testing (T1) and at follow-up
EP
(T2). The proportion of children able to delay is summarised in Figure 1.
AC C
At T1 (prior to intervention or no-intervention), there was no significant association between a child’s ability to DG and the type of training they were to undertake (nointervention or ATT), x2(1) =.58, p>.05. The calculated odds ratio (OR=1.37) revealed that despite the children in the ATT group being slightly more able to delay gratification at T1, the effect was small and not significant. A second Chi-Square test was run to explore differences in ability to DG in the nointervention and experimental conditions after the intervention or no-intervention period 11
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
(T2). This revealed that there was a significant association between ability to delay gratification and type of training (ATT or no-intervention), x2(1) =5.4, p<.05. The odds of children delaying gratification at T2 were 2.64 times higher if they had received the ATT
RI PT
intervention rather than no intervention. A McNemar’s repeated measures test was undertaken to explore improvements in
children’s ability to DG at T1 and T2 according to condition. Whilst children’s ability to delay gratification increased between T1 and T2 in both conditions, the ability to delay
SC
gratification improved significantly for children who received the ATT (p=.004). The
not statistically significant (p=.774).
M AN U
improvement between baseline and follow-up for children in the no-intervention group was
Effects of ATT on Mood and Inhibitory Control Task
Table 1 indicates that children’s mood improved from T1 to T2 in both conditions. The
TE D
mood of children in the experimental condition at T1 (M=13.56, SE=0.26) was slightly higher than children in the no-intervention condition at T1 (M=12.83, SE=0.39) and an independent
EP
t-test showed this difference was not significant t(98)=1.63, p=.11, r=.16. A 2 (Training Type: ATT vs. Control) x 2 (Time: Pre vs. Post) mixed model ANOVA was
AC C
conducted to investigate the influence of ATT on children’s mood between time points. The ANOVA revealed a significant effect of time on mood, F(1, 98)=8.362, p=.005, ηp2=.079. There was no significant interaction between training type and mood, F(1, 98)=.056 p=.813, ηp2=.001. Thus, the results showed that whether or not the child received ATT, had no significant effect on mood at T2. Next we explored effects on the day/night task. On average, at T1, children in the nointervention condition (M=11.54, SE= 0.62) performed slightly better on the day/night task 12
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
than children in the experimental condition (M=11.08, SE=0.44), but an independent t-test showed this difference was not significant t(98)=-0.61, p=0.54, r=0.06. A 2 (Training Type: ATT vs. Control) x 2 (Day/Night Score: Pre vs. Post) mixed model
RI PT
ANOVA revealed a significant effect of time on score, F(1, 98)=19.975, p<.001, ηp2=.169. There was no significant interaction between training type and day/night score, F(1, 98)=.003
task at T2. Predictors of Ability to Delay Gratification
SC
p=.960, ηp2<.001. This indicates that the ATT had no significant effect on performance on the
M AN U
The variables that predicted delay of gratification were explored using bi-variate Pearson Correlations. The following predictors correlated with delay of gratification at T2: DG at T1 (r=.47, p<.001), day/night task score at T1 (r=.37, p<.001) and at T2 (r=.40, p<.001). The relationship with mood was small and non-significant. All correlations were positive
TE D
indicating that: ability to DG at T1 was associated with ability to DG at T2; increased day/night task score (at T1 or T2) was associated with greater ability to DG at T2; and of
EP
course being in the experimental condition as previously determined. Each of the above predictors were inter-correlated with the strongest associations between
AC C
day/night score at T1 and delay of gratification at T1 (r=-.38), day/night score at T1 and day/night score at T2 (r=.64) and day/night score at T2 and delay of gratification at T1 (r=.24).
To test further the validity of the intervention effect a mixed-model logistic regression approach was taken. The outcome variable was delay of gratification (yes/no) at time 2 (T2). In the first step, to control for the differences between conditions at T1, delay of gratification at T1 was force entered into the model. To further control for the effects of age and months in 13
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
school prior to testing, these predictors were force entered at step 2. The two predictors most strongly correlated with delay of gratification at T2 were entered into the regression at step 3 using backward deletion to reduce suppressor effects (Tzelgov & Henik, 1991). These predictors were: day/night score at T1 and day/night score at T2. Finally, to test the effect of
RI PT
the ATT, the independent variable experimental condition was force entered at step 4. The final step of the regression was significant x2 = 4.71, p=.03 and the final overall model was significant, explaining 36% of the variance in DG measured at T2 (R2 =.36, x2(5) = 44.27,
SC
p<.001). Inclusion of the experimental condition as a predictor, explained an additional 3% of the variance in the outcome variable (DG at T2) when group differences at T1 in ability to
M AN U
DG, months in school, age, and day/night task score were controlled.
The final model (Table 2) indicated that delay of gratification at T1, day/night score at T2 and experimental condition were significant independent predictors. The odds ratio for condition on the final step showed that the use of ATT was associated with a unique 3.4 times
TE D
greater likelihood of delaying gratification than the no-intervention group. Discussion
EP
This study aimed to test whether Wells’ Attention Training Technique (ATT) could
AC C
improve children’s ability to delay gratification. This could be important given that childhood ability to delay gratification has consistently been shown to be a predictor of success and emotional adjustment in adult life. If the ATT creates improvement in this self-regulatory ability, it may have potential to improve longer term personal outcomes. The primary prediction was that children who had received the ATT intervention would be more able to delay gratification at follow-up testing (T2) than those in the no-intervention condition. This prediction was confirmed; children who received the ATT intervention 14
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
significantly improved in their ability to delay gratification between T1 and T2 whereas there was no significant improvement across time for those in the no-intervention condition. Children in the ATT condition were 2.64 times more likely to delay gratification at T2 compared to children in the no-intervention condition. After controlling for age, time in
RI PT
school, and individual differences in executive control, it emerged that receiving the ATT was associated with a 3.43 times greater likelihood of delaying gratification. This finding suggests the ATT intervention can be used with young children and it appeared to have a
SC
significant effect on self-regulation when this outcome was operationalised as delay of
M AN U
gratification.
A secondary question addressed whether children who performed better on the day/night task would be more able to delay gratification. The ATT did not improve performance on this task at T2. However, confirmation of this relationship was found with performance on the day/night task both at baseline testing (T1) and follow-up (T2) predicting delay of
TE D
gratification at follow-up (T2). When controlling for age, months in school, experimental condition and ability to delay gratification at T1, the relationship between day/night score
EP
became negative (i.e. for every one-point increase in day/night score, the odds of being able to delay gratification at T2 reduced by -0.36). This suggests that the control variables
AC C
moderate the relationship between day/night score and ability to delay gratification. Alternatively, given that we tested children at an age when executive functioning is developing, rather than stable, this relationship may be caused by group effects. It would therefore be advantageous to explore whether this negative relationship exists in other samples. A further secondary question we examined was whether or not children in the ATT condition would have significantly improved mood at follow-up testing (T2) compared to 15
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
those in the control condition. There was no evidence to support an effect of this kind. However, given the fact that the participants were not drawn from a clinical group, and given the difficulties inherent in assessing emotion in this group, this question may not have been
RI PT
effectively addressed in this study. In summary, given that there have been no previous attempts to use ATT in young children in order to improve self-regulatory abilities in the form of ability to delay
gratification, the significant findings of this study might be of some importance. If the ATT
SC
technique is found to improve delay of gratification as indicated by the present results it may
M AN U
do so by improving attention flexibility, leading children to be able to switch out of persistent thinking about the target. It is likely that such experiences over time will facilitate the development of metacognitions concerning competence in self-control. Such metacognitions and their interaction with attention control are a central target in metacognitive therapy (Wells, 2009) and could be extended to children with the possibility of building
TE D
psychological flexibility and resilience. The self-regulatory executive function model on which ATT is grounded (Wells and Matthews, 1994; 1996) predicts that specific attentional
EP
control techniques like ATT might provide a basis for developing adaptive metacognitions and improved self-regulation.
AC C
Future Directions and Implications Whilst all children in the present study were subject to a two-day delay between their final ATT session and the follow-up testing, it remains unclear whether the effect of the ATT is consistent and stable over a longer time period. Furthermore, the extent of the ATT was brief and we know little about the frequency of practice required by young children to maintain the effects found in this study. Whilst the Marshmallow Test has consistently been shown to provide a valid experimental paradigm for exploring children’s self-regulatory abilities, it is 16
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
difficult to assert that children’s ability within the test would be able to be transferred to different environments where they would be required to employ self-regulatory skills. Future research should therefore explore children’s self-regulatory abilities in a wider range of
RI PT
contexts. The current design compared ATT to no-intervention and therefore does not permit a
dismantling of the ATT effect. We were interested in a more basic question here; can the provision of ATT produce an effect on the marshmallow test? We do not know which
SC
components of the procedure are active, nor do we know if the practise of ATT created a
M AN U
demand effect. We were careful not to make any reference to the marshmallow task when introducing and practising the ATT and the testing of children remained separate and was not conducted by the teacher who delivered the ATT. However, further studies are required to explore precisely how the ATT achieved the effects seen in the present study for example by
TE D
comparing ATT to an active control.
Whilst a significant effect was found in respect of the primary outcome variable; delay of gratification, children’s performance on the day/night task and children’s mood, did not
EP
change as a result of the ATT intervention. However, given that many children at T1 scored within the top quartile of both the day/night task and the Faces Scale, this gave little room for
AC C
improvement at T2, creating a ceiling effect. These scales might not have been sensitive enough to capture improvement in inhibitory control and mood following ATT, and thus future work should continue to investigate potential effects of ATT in these areas. The results provide initial evidence that ATT might improve children’s self-regulatory abilities. If further exploration confirms this connection, applications of the ATT could prove beneficial in improving children’s self-regulation, and ultimately in reducing psychological distress later in life. 17
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT References
Abdel-Khalek, A. M. (2006). Measuring happiness with a single-item scale. Social Behavior and Personality, 34, 139-150.
RI PT
Ayduk, O., Medoza-Dention, R., Mischel, W., Downey, G., Peake, P. K., & Rodriguez, M. (2000). Regulating the interpersonal self: strategic self-regulation for coping with rejection sensitivity. Journal of Personality and Social Psychology, 79, 776-792.
SC
Bandura, A., & Mischel, W. (1965). Modification of self-imposed delay of reward through exposure to live and symbolic models. Journal of Personality and Social Psychology,
M AN U
2, 698-705.
Callinan, S., Johnson, D. & Wells, A. (2015). A randomised controlled study of the effects of the Attention Training Technique on traumatic stress symptoms, emotional attention
TE D
set-shifting and flexibility. Cognitive Therapy and Research, 39, 4-13. Carlson, S. M., & Moses, L. J. (2001). Individual difference in inhibitory control and Children’s Theory of Mind, Child Development, 72, 1032-1053.
EP
Casey, B. J., Somerville, L. H., Gotlib, I. H., Ayduk, O., Franklin, N. T., Askren, M. K.,
AC C
Zayas, V. (2011). Behavioral and neural correlates of delay of gratification 40 years later. Proceedings of the National Academy of Sciences, 108, 14998-15003.
Cohen, J. (1992). A Power Primer. Psychological Bulletin, 112, 155-159. Eigsti, I. M., Zayas, V., Mischel, W., Shoda, Y., Ayduk, O., Dadlani, M. B., Casey, B. J. (2006). Predicting cognitive control from preschool to late adolescence and young adulthood. Psychological Science, 17, 478-484.
18
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
Gerstadt, C. L., Hong, Y. J., & Diamond, A. (1994). The relationship between cognition and action: Performance of children 3½ - 7 years old on a Stroop-like day-night test. Cognition, 53, 129-153.
RI PT
Holder, M. D., & Coleman, B. (2008). The contribution of temperament, popularity, and physical appearance to children’s happiness. Journal of Happiness Studies, 9, 279302.
SC
Karbach, J., & Kray, J., (2009). How useful is executive control training? Age differences in near and far transfer of task-switching training. Developmental Science, 12, 978-990.
Cognition, 36, 1470-1483.
M AN U
Minear, M., & Shah, P., (2008). Training and transfer effects in task switching. Memory &
Mischel, W., & Ebbesen, E. B., (1970). Attention in delay of gratification. Journal of
TE D
Personality and Social Psychology, 16, 329-337.
Mischel, W., Shoda, Y., & Peake, P. K. (1988). The nature of adolescent competencies predicted by preschool delay of gratification. Journal of Personality and Social
EP
Psychology, 54, 687-696.
AC C
Mischel, W. Shoda, Y., & Rodriguez, M. L. (1989). Delay of gratification in children. Science, 244, 933-938.
Moffitt, T., Arseneault, L., Belsky, D., Dickson, N., Hancox, R., Harrington, H., Caspi, A. (2011).A gradient of childhood self-control predicts health, wealth, and public Safety. Proceedings of the National Academy of Sciences, 108, 2693-2698.
19
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
Nassif, Y. & Wells, A. (2014). Attention training reduces intrusive thoughts cued by a narrative of stressful life events: A controlled study. Journal of Clinical Psychology, 70, 510-517.
RI PT
Nelson, T. O., & Narens, L. (1994). Why investigate metacognition? In J. Metcalfe & A. P. Shimamura (Eds.), Metacognition: Knowing about knowing (pp. 1-25). Cambridge, MA: MIT Press.
SC
Papageorgiou, C., & Wells, A. (2000). Treatment of recurrent major depression with attention training. Cognitive and Behavioural Practice, 7, 407-413.
M AN U
Rueda, M. R., Checa, P., & Combita, L. M. (2012). Enhanced efficiency of the executive attention network after training in preschool children: immediate changes and effects after two months. Developmental Cognitive Neuroscience, 25, 192-204. Shoda, Y., Mischel, W., & Peake, P. K. (1990). Predicting adolescent cognitive and self-
TE D
regulatory competencies from pre-school delay of gratification: Identifying diagnostic condition. Developmental Psychology, 26, 978-986.
EP
Siegle, G. J., Ghinassi, F., Thase, M. E. (2007). Neurobehavioural therapies in the 21st century: Summary of the emerging field and an extended example of Cognitive
AC C
Control Training for depression. Cognitive Therapy and Research, 31, 235-262.
Streb, J., Hillie, K., Schoch, S., & Sosic-Vasic, Z. (2012). Attention training improves executive functions of preschool children. Neurology, 6, 450-454.
Spada, M.M., Caselli, G. & Wells, A. (2013). A triphasic metacognitive formulation of problem drinking. Clinical Psychology and Psychotherapy, 20, 494-500.
20
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
Tzelgov, J., & Henik, A. (1991). Suppression situations in psychological research: Definitions, implications, and applications. Psychological Bulletin, 109, 524-536. Wells, A. (1990). Panic Disorder in association with relaxation induced anxiety: An
RI PT
attentional training approach to treatment. Behavior Therapy, 21, 273-280. Wells, A. (2000). Emotional disorders & metacognition: Innovative cognitive therapy. Chichester, UK: Wiley.
SC
Wells, A. (2009). Metacognitive therapy for anxiety and depression. New York: Guilford
M AN U
Press.
Wells, A., & Matthews, G. (1994). Attention and emotion. A clinical perspective. Hove: Erlbaum.
Wells, A., & Matthews, G. (1996). Modelling cognition in emotional disorder: The S-REF
TE D
model. Behaviour Research and Therapy, 32, 867-870.
Wells, A., White, J., & Carter, K. (1997). Attention Training: Effects on anxiety and beliefs
AC C
EP
in panic and social phobia. Clinical Psychology and Psychotherapy, 4, 226-232.
21
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
T1 (SD) T2 (SD) Age 5.83 yrs (3.49 months) % able to DG 49.2 69.5 ATT Months in School 5.08 (1.01) (n=59) Emotion Score 13.56 (1.99) 14.22 (1.38) Day/Night Score 11.08 (3.39) 12.41 (2.97) Age 5.92 yrs (3.71 months) % able to DG 41.5 46.3 Control Months in School 3.71 (2.40) (n=41) Emotion Score 12.83 (2.48) 13.39 (2.51) Day/Night Score 11.54 (3.97) 12.83 (2.97) Table 1. Mean age, months in school and scores according to condition
AC C
EP
TE D
M AN U
SC
RI PT
Condition
1
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
P
Odds Ratio
95% CI Lower
95% CI Upper
DG @ T1
2.09 (0.57)
<.0005
8.09
2.67
24.52
Age
0.00 (0.00)
NS
1.00
1.00
1.01
Months Schl
-0.09 (0.16)
NS
0.91
0.67
1.24
D/N @ T2
-0.36 (0.10)
<.0005
0.70
0.57
0.85
Condition
1.23 (0.58)
0.04
3.43
1.09
10.77
Constant
-8.45 (5.04)
NS
0.00
M AN U
SC
RI PT
B (SE)
AC C
EP
TE D
Table 2. Logistic regression of predictors of children’s ability to delay gratification at time 2 [DG= delay of gratification; D/N= day/night task; Condition= ATT vs Control; T1, T2= Time 1 or 2].
2
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
Figure 1. Proportion of children able to delay gratification in each condition.
1
ACCEPTED MANUSCRIPT •
Children who received the Attention Training Technique (ATT) intervention were 2.64 times more likely to delay gratification than children who received no intervention.
•
Whether children had received the Attention Training Technique, and children’s performance on a verbal inhibition task were significant predictors of ability to delay gratification. The Attention Training Technique has potential to improve children’s self-regulatory abilities
EP
TE D
M AN U
SC
RI PT
with the implication it may reduce psychological vulnerability later in life.
AC C
•
S0005-7967(15)30058-9
DOI:
10.1016/j.brat.2015.11.009
Reference:
BRT 2936
To appear in:
Behaviour Research and Therapy
Received Date: 25 June 2015 Revised Date:
19 November 2015
Accepted Date: 19 November 2015
Please cite this article as: Murray, J., Theakston, A., Wells, A., Can the Attention Training Technique Turn One Marshmallow into Two? Improving Children’s Ability to Delay Gratification, Behaviour Research and Therapy (2015), doi: 10.1016/j.brat.2015.11.009. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
RI PT
Can the Attention Training Technique Turn One Marshmallow into Two? Improving
Joanne Murray 1
M AN U
Anna Theakston 2
SC
Children’s Ability to Delay Gratification
Adrian Wells 1
1
University of Manchester, School of Psychological Sciences, Rawnsley Building, MRI
University of Manchester, School of Psychological Sciences, Coupland 1 Building, Oxford
EP
Road, Manchester M13 9PL, UK
AC C
2
TE D
Manchester, M13 9WL UK
Corresponding author: Adrian Wells1 Email: [email protected] Tel: 0161 276 5399
1
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT Abstract
The seminal Marshmallow Test (Mischel & Ebbesen, 1970) has reliably demonstrated that children who can delay gratification are more likely to be emotionally stable and successful
RI PT
later in life. However, this is not good news for those children who can’t delay. Therefore, this study aimed to explore whether a metacognitive therapy technique, Attention Training (ATT: Wells, 1990) can improve young children’s ability to delay gratification. One hundred children participated. Classes of 5-6 year olds were randomly allocated to either the ATT or a
SC
no-intervention condition and were tested pre and post-intervention on ability to delay
M AN U
gratification, verbal inhibition (executive control), and measures of mood. The ATT intervention significantly increased (2.64 times) delay of gratification compared to the nointervention condition. After controlling for age and months in school, the ATT intervention and verbal inhibition task performance were significant independent predictors of delay of gratification. These results provide evidence that ATT can improve children’s self-regulatory
TE D
abilities with the implication that this might reduce psychological vulnerability later in life. The findings highlight the potential contribution that the Self-Regulatory Executive Function
processes.
EP
(S-REF) model could make to designing techniques to enhance children’s self-regulatory
AC C
Keywords: Children, Attention Training, Metacognition, Self-regulation, Delay of Gratification.
2
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
Can the Attention Training Technique turn one marshmallow into two? Improving children’s ability to delay gratification In order to delay gratification, people need to be able to employ self-regulation. That is,
RI PT
the ability to override and change their response to an immediate impulse or desire. Studies have shown that children who are able to self-regulate in this way are likely to experience greater success later in life (Mischel, Shoda, & Peake, 1988; Shoda, Mischel, & Peake, 1990; Ayduk et al., 2000; Eigsti et al., 2006). Conversely, inability to delay gratification in
SC
childhood is a risk factor for subsequent psychological pathology (Moffitt et al., 2011). Given
M AN U
this consistently well-supported connection, it is worthwhile to explore whether ability to delay gratification can be improved during childhood with a goal of potentially enhancing outcomes later in life. However, research has tended to overlook this extension, focusing instead on the contexts in which children are best able to delay.
TE D
Delay of gratification is a dimension of self-regulation that can be seen as a correlate of executive control (Mischel, Yuichi, & Rodriguez, 1989). In a separate line of research, studies have aimed to improve components of executive function in children (Karbach &
EP
Kray, 2008; Minear & Shah, 2008). Rueda, Checa, & Combita (2012) delivered 10 sessions of computerised attention training to healthy five year old, Spanish children in school.
AC C
Children who had received the attention training displayed faster activation of brain areas associated with executive control. Further research with four to six year old German children (Streb, Hille, Schoch, & Sosic-Vasic, 2012) delivered computerised attention training every day for one week within school. Results indicated that children who had received the intervention displayed significant improvements in inhibition and cognitive flexibility. These studies suggest that attention training strategies in young children from non-clinical samples is feasible and may be advantageous. However, these studies stand as examples of a limited 3
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
number of studies conducted in this area; they have used small samples and not specifically measured self-regulation using the marshmallow task, which we know to be a predictor of longer term outcomes.
RI PT
The attention training delivered in previous research is time consuming for a classroom setting. For example, Rueda et al’s (2012) intervention lasted seven and a half hours over five weeks. Interventions also consisted of a wide range of components intended to target various aspects of executive functioning (e.g. cognitive flexibility, inhibition, tracking, discrimination
SC
or conflict resolution). Whilst this provides a comprehensive intervention, this is challenging
M AN U
for schools to implement within their existing curriculum. Of particular importance, such multicomponent techniques are not based on any specific model linking attentional control, delay of gratification and psychological vulnerability, which is our aim. We considered how a metacognitive therapy intervention, the Attention Training
TE D
Technique (ATT; Wells, 1990) could enhance children’s ability to delay gratification. This technique is based on a specific metacognitive model, the Self-Regulatory Executive Function (S-REF) model, linking executive control to unhelpful thinking styles that cause
EP
psychological vulnerability (Wells & Matthews, 1994; 1996). Specifically, Wells and Matthews argue that psychological dysfunction is caused by perseveration of unhelpful
AC C
processing that occurs as a cognitive attentional syndrome. For example, perseveration of ideational processes is evident in addictive behaviours where repetitive thinking is characterised not only by worry and rumination but also ‘desire thinking’ (Spada, Caselli, & Wells, 2013). This thinking style prolongs internal aversive experiences leading to greater behaviours aimed at escaping anxiety, low mood or desire. The S-REF model has led to the development of treatment techniques intended to increase control over extended thinking and increase flexibility in responding to internal states. The ATT aims to enhance the capacity to 4
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
disengage perseverative processing by enhancing executive control through training individuals in externally-focused auditory attention exercises (Wells, 2000). To date, the ATT has been delivered to adults, with studies indicating its effectiveness in improving symptoms in adult patients with mood disorder (Papageorgiou & Wells, 2000), social phobia (Wells,
RI PT
White, & Carter, 1997), and traumatic stress symptoms (Callinan, Johnson, & Wells, 2015; Nassif & Wells, 2014). Significant effects have been found after 2-6 sessions (Siegle, Ghinassi, & Thase, 2007; Papageorgiou & Wells, 2000).
SC
Whilst previous studies have focused on adults experiencing psychological disorder, it is
M AN U
plausible that the ATT could disrupt perseverative processing routines that contribute to low impulse control in children. Because the ATT is designed to disconnect sustained processing and coping efforts from internal and external events it is likely that children who are trained in it could subsequently develop enhanced awareness of the independence between internal and external experiences and behaviour and increase their flexible choice over action. An
TE D
important marker for this effect in children would be an enhanced ability to delay gratification (i.e. choose not to respond to perseverative, desire related thinking). If this effect
EP
can be demonstrated then an exciting implication is that ATT may be a candidate strategy for improving psychological wellbeing (i.e. resilience) throughout life.
AC C
Aims of the Present Study
Research indicates that children begin to develop inhibitory control between 5-6 years of age (Carlson & Moses, 2001; Nelson & Narens, 1994), and therefore the primary aim of this study was to explore whether it is possible to use ATT in young children aged 5-6 years, and whether this impacts on their subsequent ability to delay gratification. It was hypothesised that those children receiving ATT would be better able to delay gratification than those in a no-intervention control condition. The testing of this hypothesis was our primary objective. 5
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
A secondary aim was exploratory and examined whether children’s ability to delay gratification was correlated with performance on a verbal Stroop task (the day/night task: Gerstadt, Hong, & Diamond, 1994), prior to and following the ATT intervention, and
RI PT
whether, in line with previous research in adults, ATT improves mood in children. Performance on the day/night task has been shown to be dependent on the development of inhibitory control; we predicted positive associations between performance on the day/night task and delay of gratification. Furthermore, we intended to use scores on this task to control
SC
for individual differences in inhibitory control when testing the effects of ATT on delay of
M AN U
gratification. Finally, we wanted to explore any effect of the ATT on children’s performance on the day/night task following the intervention.
The following specific predictions and research questions were tested: •
Primary prediction: That 5-6 year old children who receive the ATT will be more
condition.
Secondary Questions: Are children who score higher on a verbal Stroop task more
EP
•
TE D
able to delay gratification at follow-up testing than children in the no-intervention
able to delay gratification? Will children in the ATT condition have significantly
AC C
improved mood and improved performance on the day/night task compared with those in the no-intervention condition? Materials and Method
Design Within this mixed-model design, classrooms of children were randomly allocated as blocks to either the experimental condition (where children received the ATT intervention, 6
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
described below) or the no-intervention condition (where children received school activities as normal). The same seven-day study protocol was used in each school. Day one (always a Monday) involved the researcher attending school to collect baseline data. Day two to five (always Tuesday to Friday) involved the teacher administering the intervention on three
RI PT
occasions, at their convenience (or no-intervention). There was then a two day weekend
break (Saturday and Sunday) followed by the researcher returning to school on day seven (Monday) to collect follow up data.
SC
The primary dependent variable was children’s ability to delay gratification and resist a
M AN U
treat (candy), measured by the Marshmallow Test (Mischel & Ebbesen, 1970). The control/predictor variables for the main analysis were child’s age (measured in months); number of months the child had been in school; child’s mood score measured at T1 and T2 by the Faces Scale (Holder & Coleman, 2008), and; child’s ability to suppress verbal
TE D
responses, measured by the day/night task at T1 and T2 (measures are described below). Recruitment. Five classes of children were recruited from five primary schools within Greater Manchester, England. Recruitment began in summer 2012 – spring 2013 via emails
EP
to school Head Teachers. Teachers then distributed information packs and consent forms to parents. For children to be eligible to participate, they had to have parental consent and be
AC C
able to read, write and understand English. The study was approved by the University of Manchester’s Ethics Committee (Ref – 12203). Participants. One hundred children from five classes participated in the study, 59 (three classes) were randomly allocated to the experimental condition, and 41 (two classes) to the no-intervention condition. Fifty eight children were male; 42 were female. Children’s age ranged from 5.20 years to 6.52 years (M = 5.87 years, SD = 0.3 years). The time children had been in primary school at baseline testing ranged from 2 to 7 months (M = 4.52 months, SD = 7
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
1.8). One participant was withdrawn from the study due to being absent from school at T2 testing. Randomisation. Within the experimental (ATT) condition, teachers were required to
RI PT
administer the intervention to their whole class therefore there was no option to randomly assign individual children within each class to each condition. Block randomisation was used in which a whole class was randomly assigned to either the experimental or no-intervention condition. A research randomiser tool (www.randomizer.org) was used to ensure that each
SC
class had an equal chance of being allocated to the experimental or no-intervention condition.
used to determine the allocation. Apparatus
M AN U
Once a class had confirmed their participation in the study, the research randomiser tool was
Attention Training Technique. A recorded version of Wells’ (1990) Attention Training
TE D
Technique was used and standardised instructions for delivery were provided to class teachers. They were instructed to play an ATT recording on any three separate occasions during the four day intervention period. The ATT consists of approximately 11 minutes of
EP
sounds (such as bird song, traffic, clock ticking) and a voice which instructs participants to
AC C
direct attention in specific ways, for example from one sound to another and to different spatial locations.
Materials. Schools provided a room for data collection to take place undisturbed. Chairs and tables were arranged to ensure the child would be positioned in front of the treat and so that the written activities could be undertaken. For the primary outcome of the Marshmallow Test, there were two types of treat, a chocolate button or a Haribo sweet (both vegetarian) to ensure that the treat was desired by 8
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
the child. A stopwatch was used to time the 13 minutes the child had to wait alone in the room. The time started when the door was closed. A standardised score sheet was used which also allowed for recording of verbal comments made by the child during data collection.
RI PT
The following measures were also administered; to assess executive control the ‘day/night’ task was used. This consists of a standardised set of 20 day and night cards (four practice cards and 16 test cards, as described by Gerstadt et al., 1994). The dimensions of each card were 13.5 x 10 cm. The participant is instructed to say “day” when a picture of the
SC
moon is presented and “night” when the sun is presented and the number of errors are
M AN U
measured. To assess mood the Faces Scale was used (Holder & Coleman, 2008). This comprises of a five-point self-report scale and children are asked to select the face which best represents their thoughts and feelings in response to three questions: 1) “most of the time I feel;” 2) “compared to other people I feel;” 3) “right now I feel”.
TE D
Procedure
All data was collected by the researcher or a volunteer Research Assistant who had been
EP
trained in the procedure.
Baseline testing session (day one). Children’s assent was gained using a child-friendly
AC C
information sheet. Children were then asked to complete the adapted Faces Scale; the researcher provided instructions and read out the three questions individually, pausing for the child to colour-in their response. The researcher then introduced the day/night task, explaining that the child should try to answer as quickly as possible. The child was instructed to say “day” when the moon card was presented and “night” when the sun card was presented. To ensure the child understood the rule, two practice trials were undertaken (see Gerstadt et al., 1994). 9
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
To ensure results were comparable with the original study, the testing phase of the day/night task involved the presentation of 16 cards (eight day and eight night) in the random order used in the original study (Gerstadt et al., 1994): night (n), day (d), d, n, d, n, n, d, d, n,
RI PT
d, n, n, d, n, d. The cards were presented at one second intervals. The researcher then presented the child with an option of a chocolate button or a Haribo candy and placed their selection on a plate directly in front of them. The researcher told the child: “you can eat this sweet now, or if you wait until I come back, you will get two sweets”.
SC
The researcher checked the child’s understanding and left the child alone in the room, using
M AN U
the stop-watch to time 13 minutes. All distractions had been removed from the room and they were given no further direction of what to do whilst they waited. The child was positioned with their back to the door so the researcher could see them through a window in the door. If the child ate the sweet or left the room, the researcher would thank them for participating and return them to their class. If 13 minutes elapsed and the child had not eaten the sweet, the
TE D
researcher would re-enter the room and give the child another treat. Intervention stage (day two to five). Children in the no-intervention condition continued
EP
with their usual school activities between baseline and follow-up.
AC C
Children in the experimental condition also continued with their usual class activities but also received the ATT as a whole class on three separate days, chosen at the convenience of their class teacher. The ATT was delivered by either the class teacher, or a teaching assistant, using standardised instructions. Follow-up testing (day seven). The researcher returned to school seven days later to undertake follow-up testing. The Faces Scale, day/night task and the Marshmallow test were administered using the same protocol as the previous week. During the follow-up testing, the 10
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
researcher made no direct reference to the ATT to reduce demand characteristics or potential for a placebo response from the children. Results
RI PT
Pre-Study Differences In the experimental condition there were 31 males and 28 females. In the no-intervention condition there were 27 males and 14 females. Table 1 shows the mean age, months in school
SC
and scores for each condition. At T1, children in the ATT condition had been in school longer (M=5.08) on average, than children in the no-intervention condition (M=3.71). Children in
M AN U
the ATT condition were on average, slightly younger (M=5.38 years) than children in the nointervention condition (M=5.92 years). Differences in age and months in school were due to the research being conducted throughout the 2012-2013 school year.
TE D
The Effect of ATT on Delay of Gratification
Chi-Square Tests were run to explore the difference in ability to delay gratification (DG) in the no-intervention and experimental condition at baseline testing (T1) and at follow-up
EP
(T2). The proportion of children able to delay is summarised in Figure 1.
AC C
At T1 (prior to intervention or no-intervention), there was no significant association between a child’s ability to DG and the type of training they were to undertake (nointervention or ATT), x2(1) =.58, p>.05. The calculated odds ratio (OR=1.37) revealed that despite the children in the ATT group being slightly more able to delay gratification at T1, the effect was small and not significant. A second Chi-Square test was run to explore differences in ability to DG in the nointervention and experimental conditions after the intervention or no-intervention period 11
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
(T2). This revealed that there was a significant association between ability to delay gratification and type of training (ATT or no-intervention), x2(1) =5.4, p<.05. The odds of children delaying gratification at T2 were 2.64 times higher if they had received the ATT
RI PT
intervention rather than no intervention. A McNemar’s repeated measures test was undertaken to explore improvements in
children’s ability to DG at T1 and T2 according to condition. Whilst children’s ability to delay gratification increased between T1 and T2 in both conditions, the ability to delay
SC
gratification improved significantly for children who received the ATT (p=.004). The
not statistically significant (p=.774).
M AN U
improvement between baseline and follow-up for children in the no-intervention group was
Effects of ATT on Mood and Inhibitory Control Task
Table 1 indicates that children’s mood improved from T1 to T2 in both conditions. The
TE D
mood of children in the experimental condition at T1 (M=13.56, SE=0.26) was slightly higher than children in the no-intervention condition at T1 (M=12.83, SE=0.39) and an independent
EP
t-test showed this difference was not significant t(98)=1.63, p=.11, r=.16. A 2 (Training Type: ATT vs. Control) x 2 (Time: Pre vs. Post) mixed model ANOVA was
AC C
conducted to investigate the influence of ATT on children’s mood between time points. The ANOVA revealed a significant effect of time on mood, F(1, 98)=8.362, p=.005, ηp2=.079. There was no significant interaction between training type and mood, F(1, 98)=.056 p=.813, ηp2=.001. Thus, the results showed that whether or not the child received ATT, had no significant effect on mood at T2. Next we explored effects on the day/night task. On average, at T1, children in the nointervention condition (M=11.54, SE= 0.62) performed slightly better on the day/night task 12
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
than children in the experimental condition (M=11.08, SE=0.44), but an independent t-test showed this difference was not significant t(98)=-0.61, p=0.54, r=0.06. A 2 (Training Type: ATT vs. Control) x 2 (Day/Night Score: Pre vs. Post) mixed model
RI PT
ANOVA revealed a significant effect of time on score, F(1, 98)=19.975, p<.001, ηp2=.169. There was no significant interaction between training type and day/night score, F(1, 98)=.003
task at T2. Predictors of Ability to Delay Gratification
SC
p=.960, ηp2<.001. This indicates that the ATT had no significant effect on performance on the
M AN U
The variables that predicted delay of gratification were explored using bi-variate Pearson Correlations. The following predictors correlated with delay of gratification at T2: DG at T1 (r=.47, p<.001), day/night task score at T1 (r=.37, p<.001) and at T2 (r=.40, p<.001). The relationship with mood was small and non-significant. All correlations were positive
TE D
indicating that: ability to DG at T1 was associated with ability to DG at T2; increased day/night task score (at T1 or T2) was associated with greater ability to DG at T2; and of
EP
course being in the experimental condition as previously determined. Each of the above predictors were inter-correlated with the strongest associations between
AC C
day/night score at T1 and delay of gratification at T1 (r=-.38), day/night score at T1 and day/night score at T2 (r=.64) and day/night score at T2 and delay of gratification at T1 (r=.24).
To test further the validity of the intervention effect a mixed-model logistic regression approach was taken. The outcome variable was delay of gratification (yes/no) at time 2 (T2). In the first step, to control for the differences between conditions at T1, delay of gratification at T1 was force entered into the model. To further control for the effects of age and months in 13
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
school prior to testing, these predictors were force entered at step 2. The two predictors most strongly correlated with delay of gratification at T2 were entered into the regression at step 3 using backward deletion to reduce suppressor effects (Tzelgov & Henik, 1991). These predictors were: day/night score at T1 and day/night score at T2. Finally, to test the effect of
RI PT
the ATT, the independent variable experimental condition was force entered at step 4. The final step of the regression was significant x2 = 4.71, p=.03 and the final overall model was significant, explaining 36% of the variance in DG measured at T2 (R2 =.36, x2(5) = 44.27,
SC
p<.001). Inclusion of the experimental condition as a predictor, explained an additional 3% of the variance in the outcome variable (DG at T2) when group differences at T1 in ability to
M AN U
DG, months in school, age, and day/night task score were controlled.
The final model (Table 2) indicated that delay of gratification at T1, day/night score at T2 and experimental condition were significant independent predictors. The odds ratio for condition on the final step showed that the use of ATT was associated with a unique 3.4 times
TE D
greater likelihood of delaying gratification than the no-intervention group. Discussion
EP
This study aimed to test whether Wells’ Attention Training Technique (ATT) could
AC C
improve children’s ability to delay gratification. This could be important given that childhood ability to delay gratification has consistently been shown to be a predictor of success and emotional adjustment in adult life. If the ATT creates improvement in this self-regulatory ability, it may have potential to improve longer term personal outcomes. The primary prediction was that children who had received the ATT intervention would be more able to delay gratification at follow-up testing (T2) than those in the no-intervention condition. This prediction was confirmed; children who received the ATT intervention 14
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
significantly improved in their ability to delay gratification between T1 and T2 whereas there was no significant improvement across time for those in the no-intervention condition. Children in the ATT condition were 2.64 times more likely to delay gratification at T2 compared to children in the no-intervention condition. After controlling for age, time in
RI PT
school, and individual differences in executive control, it emerged that receiving the ATT was associated with a 3.43 times greater likelihood of delaying gratification. This finding suggests the ATT intervention can be used with young children and it appeared to have a
SC
significant effect on self-regulation when this outcome was operationalised as delay of
M AN U
gratification.
A secondary question addressed whether children who performed better on the day/night task would be more able to delay gratification. The ATT did not improve performance on this task at T2. However, confirmation of this relationship was found with performance on the day/night task both at baseline testing (T1) and follow-up (T2) predicting delay of
TE D
gratification at follow-up (T2). When controlling for age, months in school, experimental condition and ability to delay gratification at T1, the relationship between day/night score
EP
became negative (i.e. for every one-point increase in day/night score, the odds of being able to delay gratification at T2 reduced by -0.36). This suggests that the control variables
AC C
moderate the relationship between day/night score and ability to delay gratification. Alternatively, given that we tested children at an age when executive functioning is developing, rather than stable, this relationship may be caused by group effects. It would therefore be advantageous to explore whether this negative relationship exists in other samples. A further secondary question we examined was whether or not children in the ATT condition would have significantly improved mood at follow-up testing (T2) compared to 15
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
those in the control condition. There was no evidence to support an effect of this kind. However, given the fact that the participants were not drawn from a clinical group, and given the difficulties inherent in assessing emotion in this group, this question may not have been
RI PT
effectively addressed in this study. In summary, given that there have been no previous attempts to use ATT in young children in order to improve self-regulatory abilities in the form of ability to delay
gratification, the significant findings of this study might be of some importance. If the ATT
SC
technique is found to improve delay of gratification as indicated by the present results it may
M AN U
do so by improving attention flexibility, leading children to be able to switch out of persistent thinking about the target. It is likely that such experiences over time will facilitate the development of metacognitions concerning competence in self-control. Such metacognitions and their interaction with attention control are a central target in metacognitive therapy (Wells, 2009) and could be extended to children with the possibility of building
TE D
psychological flexibility and resilience. The self-regulatory executive function model on which ATT is grounded (Wells and Matthews, 1994; 1996) predicts that specific attentional
EP
control techniques like ATT might provide a basis for developing adaptive metacognitions and improved self-regulation.
AC C
Future Directions and Implications Whilst all children in the present study were subject to a two-day delay between their final ATT session and the follow-up testing, it remains unclear whether the effect of the ATT is consistent and stable over a longer time period. Furthermore, the extent of the ATT was brief and we know little about the frequency of practice required by young children to maintain the effects found in this study. Whilst the Marshmallow Test has consistently been shown to provide a valid experimental paradigm for exploring children’s self-regulatory abilities, it is 16
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
difficult to assert that children’s ability within the test would be able to be transferred to different environments where they would be required to employ self-regulatory skills. Future research should therefore explore children’s self-regulatory abilities in a wider range of
RI PT
contexts. The current design compared ATT to no-intervention and therefore does not permit a
dismantling of the ATT effect. We were interested in a more basic question here; can the provision of ATT produce an effect on the marshmallow test? We do not know which
SC
components of the procedure are active, nor do we know if the practise of ATT created a
M AN U
demand effect. We were careful not to make any reference to the marshmallow task when introducing and practising the ATT and the testing of children remained separate and was not conducted by the teacher who delivered the ATT. However, further studies are required to explore precisely how the ATT achieved the effects seen in the present study for example by
TE D
comparing ATT to an active control.
Whilst a significant effect was found in respect of the primary outcome variable; delay of gratification, children’s performance on the day/night task and children’s mood, did not
EP
change as a result of the ATT intervention. However, given that many children at T1 scored within the top quartile of both the day/night task and the Faces Scale, this gave little room for
AC C
improvement at T2, creating a ceiling effect. These scales might not have been sensitive enough to capture improvement in inhibitory control and mood following ATT, and thus future work should continue to investigate potential effects of ATT in these areas. The results provide initial evidence that ATT might improve children’s self-regulatory abilities. If further exploration confirms this connection, applications of the ATT could prove beneficial in improving children’s self-regulation, and ultimately in reducing psychological distress later in life. 17
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT References
Abdel-Khalek, A. M. (2006). Measuring happiness with a single-item scale. Social Behavior and Personality, 34, 139-150.
RI PT
Ayduk, O., Medoza-Dention, R., Mischel, W., Downey, G., Peake, P. K., & Rodriguez, M. (2000). Regulating the interpersonal self: strategic self-regulation for coping with rejection sensitivity. Journal of Personality and Social Psychology, 79, 776-792.
SC
Bandura, A., & Mischel, W. (1965). Modification of self-imposed delay of reward through exposure to live and symbolic models. Journal of Personality and Social Psychology,
M AN U
2, 698-705.
Callinan, S., Johnson, D. & Wells, A. (2015). A randomised controlled study of the effects of the Attention Training Technique on traumatic stress symptoms, emotional attention
TE D
set-shifting and flexibility. Cognitive Therapy and Research, 39, 4-13. Carlson, S. M., & Moses, L. J. (2001). Individual difference in inhibitory control and Children’s Theory of Mind, Child Development, 72, 1032-1053.
EP
Casey, B. J., Somerville, L. H., Gotlib, I. H., Ayduk, O., Franklin, N. T., Askren, M. K.,
AC C
Zayas, V. (2011). Behavioral and neural correlates of delay of gratification 40 years later. Proceedings of the National Academy of Sciences, 108, 14998-15003.
Cohen, J. (1992). A Power Primer. Psychological Bulletin, 112, 155-159. Eigsti, I. M., Zayas, V., Mischel, W., Shoda, Y., Ayduk, O., Dadlani, M. B., Casey, B. J. (2006). Predicting cognitive control from preschool to late adolescence and young adulthood. Psychological Science, 17, 478-484.
18
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
Gerstadt, C. L., Hong, Y. J., & Diamond, A. (1994). The relationship between cognition and action: Performance of children 3½ - 7 years old on a Stroop-like day-night test. Cognition, 53, 129-153.
RI PT
Holder, M. D., & Coleman, B. (2008). The contribution of temperament, popularity, and physical appearance to children’s happiness. Journal of Happiness Studies, 9, 279302.
SC
Karbach, J., & Kray, J., (2009). How useful is executive control training? Age differences in near and far transfer of task-switching training. Developmental Science, 12, 978-990.
Cognition, 36, 1470-1483.
M AN U
Minear, M., & Shah, P., (2008). Training and transfer effects in task switching. Memory &
Mischel, W., & Ebbesen, E. B., (1970). Attention in delay of gratification. Journal of
TE D
Personality and Social Psychology, 16, 329-337.
Mischel, W., Shoda, Y., & Peake, P. K. (1988). The nature of adolescent competencies predicted by preschool delay of gratification. Journal of Personality and Social
EP
Psychology, 54, 687-696.
AC C
Mischel, W. Shoda, Y., & Rodriguez, M. L. (1989). Delay of gratification in children. Science, 244, 933-938.
Moffitt, T., Arseneault, L., Belsky, D., Dickson, N., Hancox, R., Harrington, H., Caspi, A. (2011).A gradient of childhood self-control predicts health, wealth, and public Safety. Proceedings of the National Academy of Sciences, 108, 2693-2698.
19
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
Nassif, Y. & Wells, A. (2014). Attention training reduces intrusive thoughts cued by a narrative of stressful life events: A controlled study. Journal of Clinical Psychology, 70, 510-517.
RI PT
Nelson, T. O., & Narens, L. (1994). Why investigate metacognition? In J. Metcalfe & A. P. Shimamura (Eds.), Metacognition: Knowing about knowing (pp. 1-25). Cambridge, MA: MIT Press.
SC
Papageorgiou, C., & Wells, A. (2000). Treatment of recurrent major depression with attention training. Cognitive and Behavioural Practice, 7, 407-413.
M AN U
Rueda, M. R., Checa, P., & Combita, L. M. (2012). Enhanced efficiency of the executive attention network after training in preschool children: immediate changes and effects after two months. Developmental Cognitive Neuroscience, 25, 192-204. Shoda, Y., Mischel, W., & Peake, P. K. (1990). Predicting adolescent cognitive and self-
TE D
regulatory competencies from pre-school delay of gratification: Identifying diagnostic condition. Developmental Psychology, 26, 978-986.
EP
Siegle, G. J., Ghinassi, F., Thase, M. E. (2007). Neurobehavioural therapies in the 21st century: Summary of the emerging field and an extended example of Cognitive
AC C
Control Training for depression. Cognitive Therapy and Research, 31, 235-262.
Streb, J., Hillie, K., Schoch, S., & Sosic-Vasic, Z. (2012). Attention training improves executive functions of preschool children. Neurology, 6, 450-454.
Spada, M.M., Caselli, G. & Wells, A. (2013). A triphasic metacognitive formulation of problem drinking. Clinical Psychology and Psychotherapy, 20, 494-500.
20
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
Tzelgov, J., & Henik, A. (1991). Suppression situations in psychological research: Definitions, implications, and applications. Psychological Bulletin, 109, 524-536. Wells, A. (1990). Panic Disorder in association with relaxation induced anxiety: An
RI PT
attentional training approach to treatment. Behavior Therapy, 21, 273-280. Wells, A. (2000). Emotional disorders & metacognition: Innovative cognitive therapy. Chichester, UK: Wiley.
SC
Wells, A. (2009). Metacognitive therapy for anxiety and depression. New York: Guilford
M AN U
Press.
Wells, A., & Matthews, G. (1994). Attention and emotion. A clinical perspective. Hove: Erlbaum.
Wells, A., & Matthews, G. (1996). Modelling cognition in emotional disorder: The S-REF
TE D
model. Behaviour Research and Therapy, 32, 867-870.
Wells, A., White, J., & Carter, K. (1997). Attention Training: Effects on anxiety and beliefs
AC C
EP
in panic and social phobia. Clinical Psychology and Psychotherapy, 4, 226-232.
21
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
T1 (SD) T2 (SD) Age 5.83 yrs (3.49 months) % able to DG 49.2 69.5 ATT Months in School 5.08 (1.01) (n=59) Emotion Score 13.56 (1.99) 14.22 (1.38) Day/Night Score 11.08 (3.39) 12.41 (2.97) Age 5.92 yrs (3.71 months) % able to DG 41.5 46.3 Control Months in School 3.71 (2.40) (n=41) Emotion Score 12.83 (2.48) 13.39 (2.51) Day/Night Score 11.54 (3.97) 12.83 (2.97) Table 1. Mean age, months in school and scores according to condition
AC C
EP
TE D
M AN U
SC
RI PT
Condition
1
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
ACCEPTED MANUSCRIPT
P
Odds Ratio
95% CI Lower
95% CI Upper
DG @ T1
2.09 (0.57)
<.0005
8.09
2.67
24.52
Age
0.00 (0.00)
NS
1.00
1.00
1.01
Months Schl
-0.09 (0.16)
NS
0.91
0.67
1.24
D/N @ T2
-0.36 (0.10)
<.0005
0.70
0.57
0.85
Condition
1.23 (0.58)
0.04
3.43
1.09
10.77
Constant
-8.45 (5.04)
NS
0.00
M AN U
SC
RI PT
B (SE)
AC C
EP
TE D
Table 2. Logistic regression of predictors of children’s ability to delay gratification at time 2 [DG= delay of gratification; D/N= day/night task; Condition= ATT vs Control; T1, T2= Time 1 or 2].
2
IMPROVING CHILDREN’S ABILITY TO DELAY GRATIFICATION
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
Figure 1. Proportion of children able to delay gratification in each condition.
1
ACCEPTED MANUSCRIPT •
Children who received the Attention Training Technique (ATT) intervention were 2.64 times more likely to delay gratification than children who received no intervention.
•
Whether children had received the Attention Training Technique, and children’s performance on a verbal inhibition task were significant predictors of ability to delay gratification. The Attention Training Technique has potential to improve children’s self-regulatory abilities
EP
TE D
M AN U
SC
RI PT
with the implication it may reduce psychological vulnerability later in life.
AC C
•