108

British Journal of Health Psychology (2018), 23, 108–127

© 2017 The British Psychological Society
[Link]

A randomized controlled trial of a brief online

intervention to reduce alcohol consumption in new
university students: Combining self-affirmation,
theory of planned behaviour messages, and
implementation intentions
Paul Norman1* , David Cameron1 , Tracy Epton2 , Thomas L.
Webb1 , Peter R. Harris3 , Abigail Millings1 and Paschal
Sheeran4
1
University of Sheffield, UK
2
University of Manchester, UK
3
University of Sussex, UK
4
University of North Carolina at Chapel Hill, USA
Objectives. Excessive alcohol consumption increases when students enter university.
This study tests whether combining (1) messages that target key beliefs from the theory of
planned behaviour (TPB) that underlie binge drinking, (2) a self-affirmation manipulation
to reduce defensive processing, and (3) implementation intentions (if-then plans to avoid
binge drinking) reduces alcohol consumption in the first 6 months at university.
Design. A 2 (self-affirmation) 9 2 (TPB messages) 9 2 (implementation intention)
between-participants randomized controlled trial with 6-month follow-up.
Methods. Before starting university, students (N = 2,951) completed measures of
alcohol consumption and were randomly assigned to condition in a full-factorial design. TPB
cognitions about binge drinking were assessed immediately post-intervention (n = 2,682).
Alcohol consumption was assessed after 1 week (n = 1,885), 1 month (n = 1,389), and
6 months (n = 892) at university. TPB cognitions were assessed again at 1 and 6 months.
Results. Participants who received the TPB messages had significantly less favourable
cognitions about binge drinking (except perceived control), consumed fewer units of
alcohol, engaged in binge drinking less frequently, and had less harmful patterns of alcohol
consumption during their first 6 months at university. The other main effects were non-
significant.
Conclusions. The findings support the use of TPB-based interventions to reduce
students’ alcohol consumption, but question the use of self-affirmation and implemen-
tation intentions before starting university when the messages may not represent a threat
to self-identity and when students may have limited knowledge and experience of the
pressures to drink alcohol at university.

*Correspondence should be addressed to Paul Norman, Department of Psychology, University of Sheffield, Cathedral Court, 1
Vicar Lane, Sheffield S1 2LT, UK (email: [Link]@[Link]).
Trial registration: Current Controlled Trials ISRCTN84252967. Registered on 2 July 2014.

DOI:10.1111/bjhp.12277
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Combined alcohol intervention 109

Statement of contribution
What is already known on this subject?
 Alcohol consumption increases when young people enter university.
 Significant life transitions represent potential teachable moments to change behaviour.
 Interventions with a strong theoretical basis have been found to be more effective.

What does this study add?

 A brief online intervention delivered to students before they start university can reduce alcohol
consumption.
 The theory of planned behaviour can be used to inform the design of interventions to change health
behaviour.

Alcohol consumption, including binge drinking, increases when young people enter
university (Cameron et al., 2015; Fromme, Corbin, & Kruse, 2008), and is higher among
students than their non-student peers (Gill, 2002; Naimi et al., 2003). Binge drinking is
associated with increased risk of various negative social and health consequences,
including antisocial behaviour, physical violence, sexual assaults, unsafe sex, accidents,
and injuries (Miller, Plant, & Plant, 2005), as well as poorer academic performance
(Wechsler et al., 2002). Excessive alcohol consumption over a prolonged period of time is
associated with various long-term negative health outcomes, including cirrhosis of the
liver, heart disease, and cancer (National Health Service, 2014a). The House of Commons
Health Committee (2009, paragraph 32) has recommended that ‘universities take a much
more active role in discouraging irresponsible drinking among students’. The transition
into university represents an ideal opportunity to intervene to reduce alcohol consump-
tion in students before drinking patterns become established (Scott-Sheldon, Carey, Elliot,
Garey, & Carey, 2014).
Research indicates that health behaviour interventions that are based on theory are
more effective than those that are not (Glanz & Bishop, 2010; Noar, Benac, & Harris, 2007;
Webb, Joseph, Yardley, & Michie, 2010), as they are more likely to target the key proximal,
and modifiable, determinants of behaviour. The theory of planned behaviour (TPB; Ajzen,
1988) provides one such theoretical framework to develop interventions. According to
the TPB, intention is the proximal determinant of behaviour. Intention is, in turn,
determined by attitude (i.e., evaluations of the behaviour), subjective norms (i.e., the
perceived views of important referents), and perceived behaviour control (i.e.,
perceptions of control over, and the ease of, performing the behaviour), which is also
hypothesized to have a direct effect on behaviour. Behavioural, normative, and control
beliefs underlie each of these constructs. Recent formulations of the TPB have
distinguished between affective and cognitive attitudes, subjective and descriptive
norms, and self-efficacy and perceived control (Fishbein & Ajzen, 2010).
The TPB has been found to explain large amounts of variance in intention (44.3%) and
health behaviour (19.3%) in prospective tests (McEachan, Conner, Taylor, & Lawton,
2011), including alcohol consumption (Cooke, Dahdah, Norman, & French, 2016), and
has been applied to explain alcohol intentions and behaviour in students (Cooke,
Sniehotta, & Sch€ uz, 2007; Hagger, Anderson, Kyriakaki, & Darkings, 2007; Johnston &
White, 2003; McMillan & Conner, 2003; Norman, 2011; Norman, Armitage, & Quigley,
2007; Norman & Conner, 2006). The TPB therefore provides a strong theoretical basis for
the development of interventions to change health behaviour. In support of this idea,
Webb et al. (2010) reported that online interventions based on the TPB had a small-to-
medium-sized effect on health behaviour (d+ = 0.36), which was larger than the average
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
110 Paul Norman et al.

effect size found for all online health behaviour interventions included in their review
(d+ = 0.16).
In order to develop an intervention based on the TPB, Ajzen (1988) recommended that
researchers undertake two phases of formative research: first, to identify the modal salient
behavioural, normative, and control beliefs held by the target population and second to
assess the extent to which these beliefs are associated with intention and/or behaviour.
Accordingly, Epton, Norman, Harris, Webb, Snowsill, and Sheeran (2015) identified a
small number of beliefs that were associated with new university students’ binge drinking
intentions and behaviour including behavioural beliefs that binge drinking would be fun,
but would have a negative impact on studying, normative beliefs about the views of
friends, and control beliefs about the influence of having friends who binge drink. Epton,
Norman, et al. (2015) conducted an additional phase of formative research in which
current students were surveyed to provide arguments to target each of the chosen beliefs
that could be included in an intervention.
Interventions that attempt to change health risk behaviour can fail, however, because
recipients derogate or dismiss the health message. Leffingwell, Neumann, Leedy, and
Babitzke (2007) found that students who drank alcohol were more critical of a health
message about the risks of alcohol and rated the problem as less important than students
who did not drink alcohol. According to self-affirmation theory (Steele, 1988), such
messages may not only threaten a person’s physical integrity (by highlighting the potential
negative effect of their behaviour on their health), but also their self-integrity (i.e., their
sense of being a sensible, rational, adaptive, and morally adequate individual). People may
therefore derogate or dismiss the message in order to protect their self-integrity. Self-
affirmation, which typically involves reflecting on a cherished value or attribute in an
unrelated domain, is a simple technique that can be used to protect against threats to self-
integrity and encourage more open or unbiased processing of health messages, which
should lead to greater message acceptance and associated changes in behaviour. Epton,
Harris, Kane, van Koningsbruggen, and Sheeran (2015) reported that, on average, self-
affirmation manipulations have small but significant effects on message acceptance
(d+ = 0.17), intention (d+ = 0.14), and behaviour (d+ = 0.32). In studies with university
students, self-affirmation manipulations have been found to reduce defensive processing
of messages about the risks of alcohol (Norman & Wrona-Clarke, 2016; Scott, Brown,
Phair, Westland, & Sch€ uz, 2013) and intentions to consume alcohol (Harris & Napper,
2005; Scott et al., 2013), but not alcohol consumption (Harris & Napper, 2005; Kamboj
et al., 2016; Knight & Norman, 2016; Meier et al., 2015; Norman & Wrona-Clarke, 2016;
Scott et al., 2013).
Self-affirmation can be characterized as a motivational intervention that serves to reduce
defensive processing of health messages and promote message acceptance (Harris &
Epton, 2009). Additional volitional techniques may be required to translate good intentions
(e.g., to drink less alcohol) into behaviour (e.g., reduced alcohol consumption) (Gollwitzer
& Sheeran, 2006). Gollwitzer (1999) made the distinction between goal intentions (e.g., to
reduce alcohol consumption) and implementation intentions that specify how the goal is to
be achieved (e.g., by drinking water instead of wine at dinner). Implementation intentions
are specific if-then plans that identify a critical situation (the ‘if’ part of the plan) and link it to
an appropriate behavioural response (the ‘then’ part of the plan). Forming implementation
intentions has been found to have a medium-to-large-sized effect on health behaviour
(d+ = 0.59; Gollwitzer & Sheeran, 2006) and has been used to reduce alcohol consumption
in university students (Hagger et al., 2012; Murgraff, Abraham, & McDermot, 2007;
Murgraff, White, & Phillips, 1996; Norman & Wrona-Clarke, 2016).
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Combined alcohol intervention 111

The present research

Combining self-affirmation, messages based on the TPB, and implementation intentions
should ensure that the messages (1) are not derogated and (2) target the key beliefs
underlying students’ alcohol consumption and (3) that positive intentions are translated
into behaviour. The present study targeted students during a significant life transition, that
is, starting university, when their social and physical environments may be in a state of flux
and their beliefs more amenable to change (Heatherton & Nichols, 1994; Wood, Tam, &
Witt, 2005). Such significant life transitions represent potential ‘teachable’ moments
(Lawson & Flocke, 2009). It was hypothesized that (1) the messages based on the TPB
would lead to less favourable cognitions about binge drinking and reduced alcohol
consumption and that the effects of the messages on alcohol consumption would be
mediated by changes in cognitions about binge drinking, (2) self-affirmation would
augment the effect of the messages on cognitions about binge drinking, and (3) forming
implementation intentions would augment the effect of the messages on alcohol
consumption over the first 6 months at university.

Method
Procedure and design
Three weeks before starting university, all undergraduate students entering university
in a large UK city were sent an email inviting them to take part in the study, with a link
to the baseline questionnaire. After completing measures of demographics and alcohol
consumption, participants were randomly assigned to condition in a 2 (self-
affirmation) 9 2 (TPB-based messages) 9 2 (implementation intention) between-
participants factorial design. Thus, participants completed a self-affirmation task (i.e.,
a questionnaire about important values and attributes) or not; viewed information (i.e.,
text and videos) that targeted key beliefs about binge drinking at university or not; and
completed an implementation intentions task (i.e., if-then plans to avoid binge
drinking) or not. All participants then completed measures of TPB cognitions with
respect to binge drinking. Participants were contacted again by email 1 week,
1 month, and 6 months after starting university with a link to follow-up questionnaires
that assessed their alcohol consumption at university. They also repeated the TPB
measures at 1 and 6 months.
Participation in the study was voluntary, but was incentivized by a £100 prize draw at
each time point. Participants who completed all of the questionnaires could also win an
iPad mini. Up to three reminder emails were sent at each time point to increase response
rates. The study was approved by the Department of Psychology Research Ethics
Committee in accordance with the University’s Research Ethics Approval Procedure and
was registered with Current Controlled Trials (ISRCTN84252967).

Participants
Of the students (N = 5,832) who were sent an invitation email, 3,215 (55.1%) clicked on
the link and 2,951 (91.8%) completed the baseline demographic and alcohol consumption
measures and were randomized to condition. Of these, 2,682 (90.1%) completed the
experimental procedures and the immediate post-intervention measures. The baseline
sample comprised 1,214 males (45.3%) and 1,444 females (53.8%) (other n = 8, missing
n = 16) with a mean age of 18.76 years (SD = 1.94). Most of the sample (74.5%)
described their ethnicity as ‘White’. The sample consumed a mean of 8.16 units of alcohol
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
112 Paul Norman et al.

per week (SD = 10.91) and engaged in binge drinking a mean of 0.39 times per week
(SD = 0.73), and comprised 1,714 (64.6%) drinkers and 940 (35.4%) non-drinkers
(missing n = 28). Of the baseline sample, 1,885 (70.3%) completed a follow-up
questionnaire after 1 week, 1,389 (51.8%) after 1 month and 892 (33.2%) after 6 months
at university. Figure 1 summarizes the flow of participants through the experiment.

Interventions
Self-affirmation manipulation
The self-affirmation manipulation comprised an adapted version of the Values in Action
Strength Scale (Peterson & Seligman, 2004), as developed by Napper, Harris, and Epton
(2009). Participants rated the extent to which 32 positive traits, characteristics or qualities
(e.g., I always try to keep my word) applied to themselves on 5-point response scales
(very much like me–very much unlike me).

Messages about binge drinking

The TPB-based messages were developed on the basis of the three phases of formative
research conducted by Epton, Norman, et al. (2015). The messages targeted three key
beliefs about binge drinking, namely that engaging in binge drinking at university is fun,
that engaging in binge drinking at university has a negative impact on studies, and that
having friends who binge drink increases the likelihood of binge drinking at university.
The first message (‘You can have fun at university without binge drinking’) outlined
various ways to meet new people and have fun without binge drinking, such as joining

Eligible Participants
N = 5,832

Not Randomized
n = 2,951
No Response n = 2,617
Did Not Complete Pre-Test Q n = 264
Randomized to Condition
n = 2,951

No SAM No SAM No SAM No SAM SAM SAM SAM SAM

No Message No Message Message Message No Message No Message Message Message
No Imp Int Imp Int No Imp Int Imp Int No Imp Int Imp Int No Imp Int Imp Int
n = 369 n = 368 n = 369 n = 369 n = 368 n = 370 n = 369 n = 369

Immediate Completed Completed Completed Completed Completed Completed Completed Completed

Follow-Up n = 364 n = 303 n = 360 n = 298 n = 350 n = 345 n = 345 n = 317

One-Week Completed Completed Completed Completed Completed Completed Completed Completed

Follow-Up n = 219 n = 182 n = 218 n = 175 n = 201 n = 192 n = 207 n = 191

One-Month Completed Completed Completed Completed Completed Completed Completed Completed

Follow-Up n = 188 n = 155 n = 190 n = 153 n = 174 n = 177 n = 184 n = 168

Six-Month Completed Completed Completed Completed Completed Completed Completed Completed

Follow-Up n = 110 n = 111 n = 121 n = 105 n = 111 n = 113 n = 115 n = 106

Final Analysed Analysed Analysed Analysed Analysed Analysed Analysed Analysed

Sample n = 79 n = 75 n = 85 n = 82 n = 80 n = 80 n = 94 n = 82

Figure 1. Flow of participants through the experiment.

 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Combined alcohol intervention 113

societies (259 words). The second message (‘Binge drinking is not good for your studies’)
provided information about the impact of binge drinking on academic outcomes, and
outlined different ways by which this may occur, including missing lectures and reduced
cognitive functioning (208 words). The third message (‘Resisting social pressures to binge
drink’) highlighted the fact that most students do not binge drink on a regular basis and
that there are many reasons not to, even if friends are, including remembering that it is
‘your decision’, the financial cost of binge drinking, and being able to look after one’s
friends (216 words). Each message was followed by a brief video (c. 1 min) of students
talking about the respective issues.

Implementation intentions
Following Hagger et al. (2012), participants were asked to form up to three if-then plans
to avoid binge drinking at university. Participants were presented with brief text
highlighting the importance of making plans to avoid binge drinking at university that
included two example plans (e.g., If I feel under social pressure to binge drink, then I will
say that I have something important to do and leave). Next, participants completed a
table with text boxes for the ‘if’ and ‘then’ components of up to three plans. They were
instructed to pay particular attention to the specific situations in which the plans would be
implemented.

Measures
Alcohol consumption
At baseline, participants were asked to ‘think of a typical week and what you would
have to drink on each day of the week’. They were then presented with a table and
asked to write the type and amount of each drink that they typically consumed on each
day of the week (e.g., 1 shot of vodka, 2 pints of cider). Responses were converted into
units (= 8 g of pure alcohol) using an online calculator (National Health Service,
2014b). Both the total number of units consumed and the number of binge drinking
sessions (i.e., 8 or more units of alcohol in a single session for men and 6 or more units
for women) in a typical week were calculated and comprised the primary outcomes.
The same procedure was used to assess alcohol consumption at university, except that
at 1 week after starting university participants were asked to ‘think about what you had
to drink on each day during Intro Week’, and at 1- and 6-month follow-up, participants
were asked to think about a typical week during their first month and 6 months at
university.
At 6-month follow-up, participants also completed the 10-item Alcohol Use Disorders
Identification Test (AUDIT; Babor, Higgins-Biddle, Saunders, & Monteiro, 2001), which is
a widely used screening tool for identifying hazardous and harmful patterns of alcohol
consumption. Scores on the AUDIT can range between 0 and 40 with scores of 8 or more
being indicative of possible harmful alcohol use.

Cognitions about binge drinking

Participants completed two-item direct measures of TPB constructs, using 7-point
response scales, immediately after the intervention and 1 and 6 months after starting
university: intention (e.g., Do you intend to engage in binge drinking at university?
definitely do not–definitely do, as = .91, .90, .90), affective attitude (e.g., Engaging in
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
114 Paul Norman et al.

binge drinking at university would be. . . unpleasant–pleasant, as = .93, .93, .93),

cognitive attitude (e.g., Engaging in binge drinking at university would be. . . harmful–
beneficial, as = .83, .85, .86), subjective norms (e.g., People who are important to me
would disapprove/approve of me engaging in binge drinking at university, disap-
prove–approve, as = .76, .77, .75), descriptive norms (e.g., Most students engage in
binge drinking at university, unlikely–likely, as = .85, .84, .80), self-efficacy (e.g., If I
wanted to, engaging in binge drinking at university would be. . . difficult–easy,
as = .87, .88, .85), and perceived control (e.g., How much control do you have over
whether or not you engage in binge drinking at university? no control–complete
control, as = .69, .78, .81).
Single items assessed the extent to which participants endorsed each of the three
beliefs targeted by the messages (i.e., Engaging in binge drinking at university would be
fun, Engaging in binge drinking at university would have a negative impact on my
studies, My friends engaging in binge drinking would make my binge drinking at
university more likely) on 7-point response scales (unlikely–likely).

Results
Randomization checks
There were no significant differences between the conditions in demographics (i.e., age,
gender, ethnicity) or alcohol consumption at baseline (i.e., units consumed, frequency of
binge drinking).

Attrition analyses
Levels of attrition between randomization and completion of the immediate post-
intervention measures differed between conditions, v2(7, N = 2,951) = 149.28,
p < .001; attrition was higher among participants allocated to form implementation
intentions (14.4%) than among those who were not (3.8%). Participants lost to follow-up
were more likely to be male (80.1%) than female (71.5%), v2(1, N = 2,658) = 25.89,
p < .001, non-White (84.0%) than White (72.6%), v2(1, N = 2,676) = 36.00, p < .001,
and to consume more units of alcohol at baseline (M = 8.42, SD = 11.10) than those who
completed the follow-up questionnaires (M = 7.36, SD = 10.25), t(2652) = 2.16,
p = .03. All other comparisons were non-significant.

Alcohol consumption at university1,2,3

Two 2 (self-affirmation: yes, no) 9 2 (messages: yes, no) 9 2 (implementation intention:
yes, no) 9 3 (time: 1 week, 1 month, 6 months) mixed-measures ANCOVAs were
conducted, with units of alcohol and frequency of binge drinking assessed after 1 week,

1
Additional analyses were conducted to examine whether any of the main effects or interactions on alcohol consumption were
moderated by baseline drinker status (drinker vs. non-drinker). All of the interaction terms with drinker status were non-significant
(see Data S1).
2
Given the relatively high level of attrition, intention-to-treat analyses were also conducted for the primary outcome measures
using last observation carried forward (from 1-month follow-up) and multiple imputation methods where data were missing. The
results were broadly consistent with the original analyses (see Data S2).
3
Given that the primary outcome measures were based on count data, the data were also analysed using negative binomial
generalized linear mixed models, both with and without data imputation. Again, the results were broadly consistent with the
original analyses (see Data S3).
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Combined alcohol intervention 115

1 month, and 6 months at university as the (repeated-measures) dependent variables and

corresponding baseline measures entered as covariates. Table 1 reports the descriptive
statistics at each follow-up time point by condition.

Units of alcohol consumed

Time had a significant effect on the number of units of alcohol consumed, F(2,
1232) = 24.69, p < .001, which peaked during the first week at university and then
declined and remained stable at 1 and 6 months (see Table 2). Message condition had a
significant main effect on the number of units of alcohol consumed, F(1, 616) = 6.32,
p = .01, d = 0.20, which was lower in those who viewed the messages (see Table 2). No
other main effects or interactions were significant.

Frequency of binge drinking

Time also had a significant effect on the frequency of binge drinking, F(2, 1232) = 35.96,
p < .001, which peaked during the first week at university and then declined and
remained stable at 1 and 6 months (see Table 2). Message condition had a significant main
effect on the frequency of binge drinking, F(1, 616) = 4.25, p = .04, d = 0.17, which was
lower among participants who viewed the messages (see Table 2). The main effect of
message condition was qualified by a significant interaction with the self-affirmation
condition, F(1, 616) = 4.01, p = .046, such that the effect of message condition was
significant among non-affirmed participants (MMESS = 0.88, SE = .08; MNoMESS = 1.19,
SE = .08), F(1, 302) = 8.52, p = .004, but non-significant among self-affirmed partici-
pants (MMESS = 1.11, SE = .08; MNoMESS = 1.11, SE = .08), F(1, 317) = 0.01, p = .92.
There was also a significant three-way interaction between self-affirmation, implemen-
tation intentions, and time, F(2, 1232) = 2.38, p = .03. The interaction between self-
affirmation and time was non-significant among participants who formed implementation
intentions, F(2, 592) = 0.99, p = .32, but significant among participants who did not form
implementation intentions, F(2, 646) = 3.82, p = .02, such that self-affirmed participants
who did not form implementation intentions engaged in binge drinking more frequently
than non-affirmed participants at 1-week follow-up (MSA = 1.57, SE = .11; MNA = 1.26,
SE = .12). No other main effects or interactions were significant.

AUDIT scores
A 2 9 2 9 2 between-participants ANOVA revealed that message condition had a
significant main effect on AUDIT scores at 6-month follow-up, F(1, 875) = 4.43, p = .04,
d = 0.14, which were lower among those who viewed the messages (MMESS = 7.77,
SD = 6.21) than those who did not (MNoMESS = 8.71, SD = 6.50). No other main effects or
interactions were significant. In support of these findings, chi-square analysis revealed
that fewer participants exceeded the cut-off score for possible harmful patterns of alcohol
use in the message condition (48.1%) than in the no message condition (55.5%) at 6-month
follow-up, v2 (1, N = 882) = 4.92, p = .03.

Cognitions about binge drinking

A series of 2 (self-affirmation: yes, no) 9 2 (messages: yes, no) 9 2 (implementation
intention: yes, no) 9 3 (time: immediate, 1 month, 6 months) mixed-measures ANOVAs
 116
Paul Norman et al.

Table 1. Alcohol consumption assessed after 1 week, 1 month, and 6 months at university by condition

Non-affirmed Self-affirmed

No message Message No message Message

No II II No II II No II II No II II

Units per week

One week 23.75 (2.15) 27.25 (2.18) 19.15 (2.07) 22.12 (2.14) 24.39 (2.09) 23.46 (2.18) 23.44 (1.95) 24.72 (2.12)
One month 14.23 (1.30) 14.07 (1.32) 10.37 (1.26) 12.98 (1.30) 13.02 (1.27) 14.03 (1.32) 12.30 (1.18) 13.31 (1.29)
Six months 14.81 (1.40) 16.79 (1.42) 10.24 (1.35) 11.72 (1.39) 12.84 (1.36) 14.11 (1.42) 10.89 (1.27) 12.77 (1.38)
Binge drinking frequency
One week 1.49 (0.18) 1.74 (0.18) 1.14 (0.17) 1.32 (0.18) 1.61 (0.17) 1.35 (0.18) 1.61 (0.16) 1.66 (0.18)
One month 1.02 (0.11) 0.83 (0.11) 0.67 (0.11) 0.74 (0.11) 0.89 (0.11) 1.02 (0.11) 0.83 (0.10) 0.89 (0.11)
Six months 1.03 (0.12) 1.06 (0.12) 0.67 (0.11) 0.71 (0.11) 0.90 (0.11) 0.91 (0.12) 0.77 (0.10) 0.90 (0.11)

Notes. Values are adjusted means controlling for baseline scores. Standard errors are in parentheses.
II = implementation intention.

20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Combined alcohol intervention 117

Table 2. Alcohol consumption assessed after 1 week, 1 month, and 6 months at university by message
condition

No message Message Total

Units per week

One week 24.71 (1.07) 22.36 (1.03) 23.54 (0.75)
One month 13.84 (0.65) 12.24 (0.63) 13.04 (0.45)
Six months 14.64 (0.70) 11.41 (0.67) 13.02 (0.49)
Total 17.73 (0.69) 15.33 (0.66) 16.53 (0.48)
Binge drinking frequency
One week 1.55 (0.09) 1.43 (0.09) 1.49 (0.06)
One month 0.94 (0.06) 0.78 (0.05) 0.86 (0.04)
Six months 0.97 (0.06) 0.76 (0.06) 0.87 (0.04)
Total 1.15 (0.06) 0.99 (0.05) 1.07 (0.04)

Note. Values are adjusted means controlling for baseline scores. Standard errors are in parentheses.

was conducted, with measures of cognitions about binge drinking assessed immediately
after the intervention, and after 1 and 6 months at university as the (repeated-measures)
dependent variables. Table 3 reports the descriptive statistics at each follow-up time
point by condition.
Time had a significant effect on all cognitions (see Table 4). Intentions to binge drink,
F(2, 1330) = 10.55, p < .001, affective attitude, F(2, 1326) = 17.48, p < .001, cognitive
attitude, F(2, 1328) = 19.47, p < .001, subjective norms, F(2, 1328) = 6.02, p = .002,
descriptive norms, F(2, 1330) = 10.66, p < .001, self-efficacy, F(2, 1328) = 13.22,
p < .001, and perceived control, F(2, 1330) = 4.58, p = .01, all increased over time,
indicating more favourable cognitions about binge drinking. Post-hoc analyses revealed
that the differences between all the time points were significant, with the exception of the
difference between immediate and 1-month follow-up scores for intention and between
the 1- and 6-month follow-up scores for subjective norms and perceived control. The
belief that binge drinking at university would be fun also increased over time,
F(2, 1318) = 16.83, p < .001, whereas the belief that binge drinking would have a
negative impact on their studies decreased over time, F(2, 1312) = 9.75, p < .001.
Differences between the immediate and 1-month follow-up scores for both beliefs were
non-significant, but all other differences were significant.
Message condition had a significant main effect on all of the cognitions about binge
drinking, with the exception of perceived control (see Table 4). Participants who viewed
the messages reported weaker intentions to binge drink at university, F(1, 665) = 12.43,
p < .001, d = 0.27, less positive affective attitudes, F(1, 663) = 9.84, p = .002, d = 0.24,
less positive cognitive attitudes, F(1, 664) = 12.69, p < .001, d = 0.28, lower subjective
norms, F(1, 664) = 8.22, p = .004, d = 0.22, lower descriptive norms, F(1, 665) = 53.29,
p < .001, d = 0.56, and lower self-efficacy, F(1, 664) = 5.38, p = .02, d = 0.18, than
participants who did not view the messages. Participants who viewed the messages were also
less likely to believe that binge drinking at university would be fun, F(1, 659) = 8.17, p = .04,
d = 0.22, and more likely to believe that it would have a negative impact on their studies, F(1,
656) = 26.19, p < .001, d = 0.40, than participants who did not view the messages.
The significant effects of message condition on intention, F(2, 1330) = 3.09, p = .046,
affective attitude, F(2, 1326) = 5.45, p = .004, cognitive attitude, F(2, 1328) = 7.41,
p = .001, subjective norms, F(2, 1328) = 3.39, p = .03, descriptive norms, F(2,
1330) = 11.72, p < .001, and the belief that binge drinking would impact on studies,
 118
Table 3. Theory of planned behaviour measures assessed immediately post-intervention and after 1 month and 6 months at university by condition

Non-affirmed Self-affirmed
No message Message No message Message

No II II No II II No II II No II II
Paul Norman et al.

Intention
Immediate 3.77 (0.20) 3.44 (0.22) 2.60 (0.20) 2.96 (0.22) 3.52 (0.21) 3.21 (0.22) 3.00 (0.20) 2.99 (0.21)
One month 3.84 (0.22) 3.53 (0.23) 2.82 (0.21) 3.22 (0.23) 3.53 (0.22) 3.20 (0.24) 2.96 (0.22) 3.12 (0.22)
Six months 3.65 (0.23) 3.60 (0.24) 2.80 (0.22) 3.47 (0.24) 3.74 (0.23) 3.57 (0.25) 3.12 (0.23) 3.31 (0.24)
Affective attitude
Immediate 3.63 (0.19) 3.27 (0.20) 2.67 (0.18) 2.71 (0.20) 3.37 (0.19) 3.27 (0.20) 2.85 (0.19) 2.98 (0.19)
One month 3.74 (0.20) 3.33 (0.21) 2.93 (0.19) 3.01 (0.21) 3.35 (0.20) 3.13 (0.21) 2.87 (0.20) 3.23 (0.20)
Six months 3.65 (0.21) 3.49 (0.22) 3.06 (0.20) 3.35 (0.22) 3.48 (0.21) 3.40 (0.22) 3.03 (0.21) 3.40 (0.21)
Cognitive attitude
Immediate 2.63 (0.12) 2.44 (0.13) 2.09 (0.12) 1.94 (0.13) 2.43 (0.13) 2.51 (0.13) 1.98 (0.12) 2.17 (0.13)
One month 2.73 (0.13) 2.46 (0.14) 2.20 (0.12) 2.20 (0.14) 2.50 (0.13) 2.39 (0.14) 2.17 (0.13) 2.41 (0.13)
Six months 2.72 (0.14) 2.65 (0.15) 2.34 (0.14) 2.37 (0.15) 2.50 (0.14) 2.48 (0.15) 2.25 (0.14) 2.69 (0.14)
Subjective norm
Immediate 3.27 (0.16) 2.90 (0.17) 2.57 (0.16) 2.56 (0.13) 2.75 (0.17) 3.11 (0.18) 2.49 (0.16) 2.81 (0.17)
One month 3.25 (0.15) 3.06 (0.16) 2.76 (0.15) 2.55 (0.14) 2.93 (0.15) 3.03 (0.16) 2.76 (0.16) 2.98 (0.16)
Six months 3.11 (0.16) 3.04 (0.17) 2.67 (0.16) 2.74 (0.15) 2.87 (0.16) 3.17 (0.17) 2.98 (0.16) 3.18 (0.16)
Descriptive norm
Immediate 5.35 (0.13) 4.91 (0.13) 4.31 (0.12) 4.37 (0.13) 5.26 (0.13) 5.20 (0.14) 4.48 (0.12) 4.52 (0.13)
One month 5.15 (0.12) 5.10 (0.12) 4.65 (0.11) 4.66 (0.15) 5.02 (0.12) 5.16 (0.13) 4.80 (0.11) 4.68 (0.12)
Six months 5.18 (0.12) 5.15 (0.13) 4.69 (0.11) 4.70 (0.13) 5.23 (0.12) 5.26 (0.13) 4.71 (0.12) 4.95 (0.12)
Self-efficacy
Immediate 5.52 (0.20) 5.36 (0.21) 4.87 (0.20) 5.18 (0.21) 5.42 (0.20) 5.14 (0.22) 4.76 (0.20) 5.24 (0.21)
One month 6.13 (0.17) 5.67 (0.18) 5.61 (0.16) 5.49 (0.18) 5.96 (0.17) 5.68 (0.18) 5.48 (0.17) 5.91 (0.17)
Six months 6.10 (0.16) 5.94 (0.17) 5.88 (0.16) 5.10 (0.17) 6.12 (0.17) 5.95 (0.18) 5.56 (0.16) 6.10 (0.17)

Continued

20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
 Table 3. (Continued)

Non-affirmed Self-affirmed

No message Message No message Message

No II II No II II No II II No II II

Perceived control
Immediate 6.15 (0.12) 6.15 (0.12) 6.32 (0.11) 6.06 (0.12) 6.08 (0.12) 6.19 (0.13) 6.31 (0.11) 6.20 (0.12)
One month 6.19 (0.12) 6.17 (0.12) 6.36 (0.11) 6.17 (0.12) 6.37 (0.12) 6.25 (0.13) 6.30 (0.12) 6.39 (0.12)
Six months 6.18 (0.12) 6.37 (0.12) 6.37 (0.11) 6.33 (0.12) 6.09 (0.12) 6.18 (0.13) 6.50 (0.11) 6.46 (0.12)
Belief 1 – Fun
Immediate 4.03 (0.21) 3.65 (0.22) 3.18 (0.20) 3.03 (0.22) 3.67 (0.21) 3.82 (0.23) 3.14 (0.21) 3.50 (0.22)
One month 4.21 (0.22) 3.72 (0.23) 3.45 (0.21) 3.40 (0.23) 3.91 (0.22) 3.74 (0.24) 3.43 (0.21) 3.85 (0.23)
Six months 4.06 (0.22) 3.92 (0.24) 3.47 (0.22) 3.56 (0.24) 3.85 (0.23) 4.04 (0.24) 3.58 (0.22) 4.04 (0.23)
Belief 2 – Impact on studies
Immediate 5.26 (0.16) 5.40 (0.17) 6.11 (0.15) 6.36 (0.16) 5.67 (0.16) 5.37 (0.17) 6.27 (0.15) 5.93 (0.16)
One month 5.19 (0.16) 5.39 (0.16) 6.04 (0.15) 5.96 (0.16) 5.65 (0.16) 5.34 (0.17) 5.96 (0.15) 5.86 (0.16)
Six months 5.23 (0.17) 5.39 (0.18) 5.67 (0.16) 5.82 (0.18) 5.79 (0.17) 5.43 (0.18) 5.69 (0.17) 5.43 (0.17)
Belief 3 – Friends bingeing
Immediate 4.25 (0.22) 3.77 (0.23) 3.57 (0.21) 3.91 (0.23) 3.91 (0.22) 3.67 (0.24) 3.76 (0.21) 3.62 (0.22)
One month 4.42 (0.22) 3.78 (0.23) 3.60 (0.21) 3.91 (0.23) 3.93 (0.22) 3.69 (0.24) 3.81 (0.22) 3.90 (0.23)
Six months 4.24 (0.23) 3.99 (0.24) 3.66 (0.22) 4.18 (0.24) 4.06 (0.23) 4.03 (0.25) 3.81 (0.23) 3.74 (0.24)

Notes. Values are means. Standard errors are in parentheses.

II = implementation intention.
Combined alcohol intervention
119

20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
120 Paul Norman et al.

Table 4. Theory of planned behaviour measures assessed immediately post-intervention and after
1 month and 6 months at university by message condition

No message Message Total

Intention
Immediate 3.48 (0.11) 2.89 (0.10) 3.19 (0.07)
One month 3.52 (0.11) 3.03 (0.11) 3.28 (0.07)
Six months 3.64 (0.12) 3.17 (0.12) 3.41 (0.08)
Total 3.55 (0.11) 3.03 (0.10) 3.29 (0.07)
Affective attitude
Immediate 3.38 (0.10) 2.80 (0.09) 3.09 (0.07)
One month 3.39 (0.10) 3.01 (0.11) 3.20 (0.07)
Six months 3.50 (0.11) 3.21 (0.11) 3.36 (0.08)
Total 3.43 (0.10) 3.01 (0.09) 3.21 (0.07)
Cognitive attitude
Immediate 2.50 (0.06) 2.04 (0.06) 2.27 (0.05)
One month 2.52 (0.07) 2.24 (0.06) 2.38 (0.05)
Six months 2.59 (0.07) 2.41 (0.07) 2.50 (0.05)
Total 2.53 (0.06 2.23 (0.06) 2.38 (0.04)
Subjective norm
Immediate 3.01 (0.08) 2.61 (0.08) 2.81 (0.06)
One month 3.07 (0.08) 2.76 (0.08) 2.92 (0.06)
Six months 3.05 (0.08) 2.89 (0.08) 2.97 (0.06)
Total 3.04 (0.07) 2.75 (0.07) 2.90 (0.05)
Descriptive norm
Immediate 5.18 (0.07) 4.92 (0.06) 4.80 (0.05)
One month 5.11 (0.06) 4.70 (0.06) 4.90 (0.04)
Six months 5.21 (0.06) 4.77 (0.06) 5.00 (0.04)
Total 5.16 (0.05) 4.63 (0.05) 4.90 (0.04)
Self-efficacy
Immediate 5.36 (0.10) 5.01 (0.10) 5.19 (0.07)
One month 5.86 (0.09) 5.63 (0.09) 5.74 (0.06)
Six months 6.03 (0.09) 5.81 (0.08) 5.92 (0.06)
Total 5.75 (0.08) 5.48 (0.08) 5.61 (0.06)
Perceived control
Immediate 6.14 (0.06) 6.22 (0.06) 6.18 (0.04)
One month 6.25 (0.06) 6.31 (0.06) 6.28 (0.04)
Six months 6.20 (0.06) 6.42 (0.06) 6.31 (0.04)
Total 6.20 (0.05) 6.32 (0.05) 6.26 (0.03)
Belief 1 – Fun
Immediate 3.80 (0.11) 3.21 (0.11) 3.30 (0.08)
One month 3.89 (0.12) 3.53 (0.11) 3.71 (0.08)
Six months 3.97 (0.12) 3.66 (0.11) 3.82 (0.08)
Total 3.89 (0.10) 3.47 (0.10) 3.68 (0.07)
Belief 2 – Impact on studies
Immediate 5.43 (0.08) 6.17 (0.08) 5.76 (0.06)
One month 5.39 (0.08) 5.96 (0.08) 5.67 (0.06)
Six months 5.46 (0.09) 5.65 (0.09) 5.56 (0.06)
Total 5.43 (0.07) 5.92 (0.07) 5.68 (0.07)

Continued
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Combined alcohol intervention 121

Table 4. (Continued)

No message Message Total

Belief 3 – Friends bingeing

Immediate 3.90 (0.11) 3.72 (0.11) 3.81 (0.08)
One month 3.95 (0.11) 3.81 (0.11) 3.88 (0.08)
Six months 4.08 (0.12) 3.85 (0.12) 3.96 (0.08)
Total 3.89 (0.10) 3.79 (0.10) 3.88 (0.07)

Note. Values are means. Standard errors are in parentheses.

F(2, 1312) = 13.61, p < .001, were qualified by significant interactions with time.
Post-hoc analyses indicated that the effects of the messages weakened over time, although
the effects on intention, affective attitude, and descriptive norms remained significant at
6-month follow-up. No other main effects or interactions were significant.

Mediation analyses
Mediation analyses assessed whether the effects of the message condition on alcohol
consumption were mediated by changes in cognitions about binge drinking (Preacher &
Hayes, 2008). Message condition was entered as an independent variable along with the
measures of the TPB assessed immediately post-intervention as potential mediators and
alcohol consumption at baseline as a covariate. Alcohol consumption at 6-month follow-
up was the dependent variable.
The direct effect of message condition on the number of units of alcohol consumed at
6-month follow-up, B = 3.12, SE = .83, p < .001, was reduced to non-significance
when the TPB variables were controlled, B = 1.52, SE = .83, p = .07, thereby
suggesting mediation. Using bootstrapping procedures, the total indirect effect was
found to be significant, B = 1.50, SE = .42, CI = 2.39 to 0.75. Only the individual
indirect effects via self-efficacy, B = 0.19, SE = .10, CI = 0.45 to 0.05, and intention,
B = 1.10, SE = .38, CI = 2.11 to 0.52, were significant.
The direct effect of the message condition on the frequency of binge drinking at 6-
month follow-up, B = 0.20, SE = .07, p = .004, was reduced to non-significance when
the TPB variables were controlled, B = 0.06, SE = .07, p = .37. Using bootstrapping
procedures, the total indirect effect was found to be significant, B = 0.14, SE = .04,
CI = 0.21 to 0.07. Again, only the individual indirect effects via self-efficacy,
B = 0.02, SE = .01, CI = 0.05 to 0.01, and intention, B = 0.08, SE = .03,
CI = 0.15 to 0.04, were significant.

Discussion
The present study employed a full-factorial design to test the effect of combining self-
affirmation, messages based on the TPB and implementation intentions on alcohol
consumption in new university students. The messages had significant effects on the
quantity of alcohol consumed, reducing the frequency of binge drinking and harmful
patterns of alcohol use over students’ first 6 months at university. Moreover, these effects
of the messages on alcohol consumption did not diminish over time. The messages also
had significant effects on (reducing) intentions to binge drink, cognitive attitudes,
subjective norms, descriptive norms, and self-efficacy, although some of these effects
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
122 Paul Norman et al.

weakened over time. The effects of the messages on both the quantity of alcohol
consumed and the frequency of binge drinking were mediated by TPB variables with
significant indirect effects through intention and self-efficacy which, according to the
TPB, are proximal determinants of behaviour.
The effect sizes for the TPB-based messages on the quantity of alcohol consumed
(d = 0.20) and the frequency of binge drinking (d = 0.17), although small, are larger than the
average effect sizes reported by interventions targeting alcohol consumption in first-year
university students (d+s = 0.13, 0.07; Scott-Sheldon et al., 2014) and for online alcohol
interventions (d+s = 0.15, 0.07; Black, Mullan, & Sharpe, 2016). Furthermore, the finding that
the effects of the messages were mediated by cognitions about binge drinking (i.e., intention
and self-efficacy) provides strong support for the TPB (Norman & Conner, 2015). The
significant effects of the messages contrast with the non-significant effects found for self-
affirmation and implementation intentions. This may indicate that the messages were (1)
sufficiently relevant to students not to be dismissed or derogated and (2) sufficiently
persuasive to produce changes in behaviour without the need to form if-then plans. Extensive
formative research was conducted to identify the key beliefs underlying binge drinking at
university and to develop of messages to target them (Epton, Norman, et al., 2015), which is
likely to have increased their relevance and effectiveness. In addition, the messages were
presented to students just before a significant life transition when their beliefs may have been
more malleable (Heatherton & Nichols, 1994; Lawson & Flocke, 2009).
Non-significant effects of self-affirmation on measures of message acceptance (Kamboj
et al., 2016; Knight & Norman, 2016; Meier et al., 2015) and alcohol consumption (Harris
& Napper, 2005; Kamboj et al., 2016; Knight & Norman, 2016; Meier et al., 2015; Norman
& Wrona-Clarke, 2016; Scott et al., 2013) have been reported in other studies with
university students, although significant effects have been reported with retail workers
(Armitage, Harris, & Arden, 2011). The non-significant effects of self-affirmation in the
present study may, in part, be due to the fact that participants completed the self-
affirmation manipulation before they entered university. It is possible that the messages
were not perceived as a threat to self-integrity given that they targeted a future, rather than
a current, behaviour (i.e., binge drinking at university). As a result, there may have been
little need for participants to self-affirm to overcome defensive processing. The fact that
the messages were found to have significant effects on cognitions about binge drinking
and subsequent alcohol consumption at university is consistent with such an explanation.
Future research could therefore test the effects of repeated administration of self-
affirmation manipulations at different points across the transition into university when
messages about the risks of binge drinking may be more threatening.
The present research also found some evidence that the self-affirmation manipulation
may have been counterproductive, such that the messages only reduced the frequency of
binge drinking at university if participants did not self-affirm. Knight and Norman (2016)
argued that self-affirmation manipulations may inadvertently prime social goals that are
closely associated with drinking in university students, thereby counteracting the effects
of such manipulations on the processing of health risk information about alcohol.
Consistent with this argument, Norman and Wrona-Clarke (2016) found that university
students who affirmed a social value had stronger intentions to engage in binge drinking
than those who affirmed a non-social value. Similarly, Voisin, Girandola, David, and Aim
(2016) found that a self-affirmation manipulation only reduced students’ derogation of a
health message about the risks of binge drinking when the message did not contain
incongruent normative information about the prevalence of binge drinking in young
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Combined alcohol intervention 123

people. Alternative self-affirmation manipulations that avoid priming social goals may
need to be developed.
The non-significant effects of forming implementation intentions contrast with the
significant effects on alcohol consumption reported in other studies (Hagger et al., 2012;
Murgraff et al., 1996, 2007; Norman & Wrona-Clarke, 2016). One key difference between
the present study and previous work is that implementation intentions were formed
before students started university. Research on the hot–cold empathy gap (Loewenstein,
1996) suggests that people often fail to appreciate in advance how ‘hot’ affective states
(e.g., feelings of excitement) will influence their behaviour at the moment of acting.
Furthermore, Sugarman and Carey (2009) have argued that experienced drinkers at
university are likely to have developed appropriate protective strategies. In contrast,
incoming students may lack knowledge of the high-risk situations they are likely to
encounter at university and how to deal with them. As a result, they may make poor-
quality plans to avoid binge drinking at university. The task of forming implementation
intentions could therefore be delayed or repeated at university when students have more
experience of drinking contexts and pressures as occurs in the AlcoholEdu for College
programme (EverFi, 2016), which is used in many US universities.
The present study had a number of limitations that should be noted. First, participants
randomly allocated to the implementation intentions condition were more likely to drop
out of the study between randomization and completion of the immediate post-
intervention measures. This may simply reflect the additional burden of this task or that
participants did not consider making plans to avoid binge drinking before starting
university to be relevant. Second, participants lost to follow-up were more likely to be
male and non-White and consumed more alcohol at baseline than those who completed all
follow-up questionnaires, thereby limiting the generalizability of the findings. Third,
attrition across the follow-up period was relatively high. Intention-to-treat analyses were
therefore conducted to examine the effect of the interventions on the primary outcomes
(i.e., units of alcohol consumed and frequency of binge drinking) at 6-month follow-up
using both last observation carried forward (from 1-month follow-up) and multiple
imputation methods. These additional analyses produced broadly consistent results.
Fourth, alcohol consumption was assessed by self-report which may introduce self-
presentation biases. However, self-report measures of the type used in the present study
have been found to provide accurate estimates of alcohol consumption (Del Boca & Noll,
2000). Finally, the present study tested a single TPB intervention. Testing separate
manipulations of attitudes, norms, and perceptions of control in a full-factorial design
would provide a stronger experimental test of the TPB (Sniehotta, Presseau, & Ara ujo-
Soares, 2014).
A number of important implications can be drawn from the current findings. First, the
findings support the idea that brief interventions delivered to students before they enter
university can reduce alcohol consumption at university (e.g., Hustad, Barnett, Borsari, &
Jackson, 2010). Second, the findings suggest that the TPB provides a strong theoretical
framework for developing interventions to reduce alcohol consumption in university
students. Third, the findings suggest that the timing of interventions may influence their
effectiveness. The significant effects of the messages may, in part, be due to the fact that
they were delivered just before a significant life transition (i.e., at a ‘teachable moment’)
when students’ beliefs about binge drinking may have been more amenable to change
(Lawson & Flocke, 2009). In contrast, administering a self-affirmation manipulation at this
point in time may be unnecessary given that the message targeted a future, rather than a
current, behaviour (i.e., binge drinking at university). As a result, the messages may not
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
124 Paul Norman et al.

have represented a threat to participants’ self-integrity. Similarly, forming implementation

intentions to avoid binge drinking before starting university may not be effective, as
incoming students may not have sufficient knowledge of the high-risk situations they are
likely to encounter at university. Thus, it may be premature to suggest that prompting
students to self-affirm or form implementation intentions (or do both) is not an effective
way to reduce alcohol consumption. Instead, future interventions to reduce binge
drinking in new university students might test the effects of repeated (or delayed)
administration of different intervention components across the transition into university
(e.g., before and after students have entered university).

Conflict of Interest
All authors declare no conflict of interest.

References
Ajzen, I. (1988). Attitudes, personality and behavior. Milton Keynes, UK: Open University Press.
Armitage, C. J., Harris, P. R., & Arden, M. A. (2011). Evidence that self-affirmation reduces alcohol
consumption: Randomized exploratory trial with a new, brief means of self-affirming. Health
Psychology, 30, 633–641. [Link]
Babor, T. F., Higgins-Biddle, J. C., Saunders, J. B., & Monteiro, M. G. (2001). The alcohol use disorders
identification test. Geneva, Switzerland: World Health Organization.
Black, N., Mullan, B., & Sharpe, L. (2016). Computer-delivered interventions for reducing alcohol
consumption: Meta-analysis and meta-regression using behaviour change techniques and
theory. Health Psychology Review, 10, 341–357. [Link]
1168268
Cameron, D., Epton, T., Norman, P., Sheeran, P., Harris, P. R., Webb, T. L., . . . Shah, I. (2015).
A theory-based online health behaviour intervention for new university students (U@Uni:
LifeGuide): Results from a repeat randomised controlled trial. BMC Trials, 16, 555. [Link]
org/10.1186/s13063-015-1092-4
Cooke, R., Dahdah, M., Norman, P., & French, D. (2016). How well does the theory of planned
behaviour predict alcohol consumption? A systematic review and meta-analysis. Health
Psychology Review, 10, 128–167. [Link]
Cooke, R., Sniehotta, F., & Sch€ uz, B. (2007). Predicting binge-drinking behaviour using an extended
TPB: Examining the impact of anticipated regret and descriptive norms. Alcohol and
Alcoholism, 42, 84–91. [Link]
Del Boca, F. K., & Noll, J. A. (2000). Truth or consequences: The validity of self-report data in health
services research on addictions. Addiction, 95(Suppl. 3), S347–S360.
Epton, T., Harris, P. R., Kane, R., van Koningsbruggen, G. M., & Sheeran, P. (2015). The impact of
self-affirmation on health-behavior change: A meta-analysis. Health Psychology, 34, 187–196.
[Link]
Epton, T., Norman, P., Harris, P., Webb, T., Snowsill, A., & Sheeran, P. (2015). Development of
theory-based health messages: Three-phase programme of formative research. Health
Promotion International, 30, 756–768. [Link]
EverFi (2016). EduAlcohol for college. Retrieved from [Link]
oholedu/
Fishbein, M., & Ajzen, I. (2010). Predicting and changing behavior: The reasoned action
approach. New York, NY: Psychology Press.
Fromme, K., Corbin, W. R., & Kruse, M. I. (2008). Behavioral risks during the transition from high
school to college. Developmental Psychology, 44, 1497–1504. [Link]
a0012614
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Combined alcohol intervention 125

Gill, J. S. (2002). Reported levels of alcohol consumption and binge drinking within the UK
undergraduate student population over the last 25 years. Alcohol and Alcoholism, 37, 109–120.
[Link]
Glanz, K., & Bishop, D. B. (2010). The role of behavioral science theory in development and
implementation of public health interventions. Annual Review of Public Health, 31, 399–418.
[Link]
Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple plans. American
Psychologist, 29, 971–979. [Link]
Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achievement: A meta-
analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69–119.
[Link]
Hagger, M., Anderson, M., Kyriakaki, M., & Darkings, S. (2007). Aspects of identity and their
influence on intentional behavior: Comparing effects for three health behaviors. Personality
and Individual Differences, 42, 355–367. [Link]
Hagger, M. S., Lonsdale, A., Koka, A., Hein, V., Pasi, H., Lintunen, T., & Chatzisarantis, N. L. D. (2012).
An intervention to reduce alcohol consumption in undergraduate students using
implementation intentions and mental simulations: A cross-national study. International
Journal of Behavioral Medicine, 19, 82–96. [Link]
Harris, P. R., & Epton, T. (2009). The impact of self-affirmation on health cognition, health behaviour
and other health-related responses: A narrative review. Social and Personality Psychology
Compass, 3, 962–978. [Link]
Harris, P. R., & Napper, L. (2005). Self-affirmation and the biased processing of threatening health-
risk information. Personality and Social Psychology Bulletin, 31, 1250–1263. [Link]
10.1177/014167205274694
Heatherton, T. F., & Nichols, P. A. (1994). Personal accounts of successful versus failed attempts at
life change. Personality and Social Psychology Bulletin, 20, 664–675. [Link]
0146167294206005
House of Commons Health Committee. (2009). Alcohol. Retrieved from [Link]
[Link]/pa/cm200910/cmselect/cmhealth/151/[Link]
Hustad, J. T. P., Barnett, N. P., Borsari, B., & Jackson, M. J. (2010). Web-based alcohol prevention for
incoming college students: A randomized controlled trial. Addictive Behaviors, 35, 183–189.
[Link]
Johnston, K. L., & White, K. M. (2003). Binge drinking: A test of the role of group norms in the theory
of planned behaviour. Psychology and Health, 18, 63–77. [Link]
0887044021000037835
Kamboj, S. K., Place, H., Barton, J. A., Linke, S., Curran, H. V., & Harris, P. R. (2016). Processing of
alcohol-related health threat in at-risk drinkers: An online study of gender-related self-affirmation
effects. Alcohol and Alcoholism, 51, 756–762. [Link]
Knight, R., & Norman, P. (2016). Impact of brief self-affirmation manipulations on university
students’ reactions to risk information about binge drinking. British Journal of Health
Psychology, 21, 570–583. [Link]
Lawson, P. J., & Flocke, S. A. (2009). Teachable moments for health behavior change: A concept
analysis. Patient Education and Counseling, 76, 25–30. [Link]
11.002
Leffingwell, T. R., Neumann, C., Leedy, M. J., & Babitzke, A. C. (2007). Defensively biased responding
to risk information among alcohol-using college students. Addictive Behaviors, 32, 158–165.
[Link]
Loewenstein, G. (1996). Out of control: Visceral influences on behavior. Organizational Behavior
and Human Decision Processes, 65, 272–292. [Link]
McEachan, R. R. C., Conner, M., Taylor, N. J., & Lawton, R. J. (2011). Prospective prediction of
health-related behaviours with the Theory of Planned Behaviour: A meta-analysis. Health
Psychology Review, 5, 97–144. [Link]
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
126 Paul Norman et al.

McMillan, B., & Conner, M. (2003). Using the theory of planned behaviour to understand alcohol and
tobacco use in students. Psychology, Health and Medicine, 8, 317–328. [Link]
1080/1354850031000135759
Meier, E., Miller, M. B., Lechner, W. V., Lombardi, N., Claborn, K. R., & Leffingwell, T. R. (2015). The
inability of self-affirmations to decrease defensive bias toward an alcohol-related risk message
among high-risk college students. Journal of American College Health, 63, 324–329. https://
[Link]/10.1080/07448481.2014.1003377
Miller, P., Plant, M., & Plant, M. (2005). Spreading out or concentrating weekly consumption:
Alcohol problems and other consequences within a UK population sample. Alcohol and
Alcoholism, 40, 461–468. [Link]
Murgraff, V., Abraham, C., & McDermot, M. (2007). Reducing Friday alcohol consumption among
moderate, women drinkers: Evaluation of a brief evidence-based intervention. Alcohol and
Alcoholism, 42, 37–41. [Link]
Murgraff, V., White, D., & Phillips, K. (1996). Moderating binge drinking: It is possible to change
behaviour if you plan it in advance. Alcohol and Alcoholism, 31, 577–582. [Link]
1093/[Link].a008193
Naimi, T. S., Brewer, R. D., Mokdad, A., Denny, C., Sredula, M. K., & Marks, J. S. (2003). Binge
drinking among U.S. adults. Journal of the American Medical Association, 289, 70–75.
[Link]
Napper, L., Harris, P. R., & Epton, T. (2009). Developing and testing a self-affirmation manipulation.
Self and Identity, 8, 45–62. [Link]
National Health Service. (2014a). Binge drinking. Retrieved from [Link]
ohol/Pages/[Link]
National Health Service. (2014b). Alcohol units. Retrieved from [Link]
ohol/Pages/[Link]
Noar, S. M., Benac, C. N., & Harris, M. S. (2007). Does tailoring matter? Meta-analytic review of
tailored print health Behavior change interventions. Psychological Bulletin, 133, 673–693.
[Link]
Norman, P. (2011). The theory of planned behaviour and binge drinking among undergraduate
students: Assessing the impact of habit strength. Addictive Behaviors, 36, 502–507. [Link]
org/10.1016/[Link].2011.01.025
Norman, P., Armitage, C. J., & Quigley, C. (2007). The theory of planned behavior and binge
drinking: Assessing the impact of binge drinker prototypes. Addictive Behaviors, 32, 1753–
1768. [Link]
Norman, P., & Conner, M. (2006). The theory of planned behaviour and binge drinking: Assessing
the moderating role of past behaviour within the theory of planned behaviour. British Journal of
Health Psychology, 11, 55–70. [Link]
Norman, P., & Conner, M. (2015). Predicting and changing health behaviour: Future research
directions. In M. Conner & P. Norman (Eds.), Predicting and changing health behaviour:
Research and practice with social cognition models (3rd ed., pp. 390–430). Buckingham, UK:
Open University Press.
Norman, P., & Wrona-Clarke, A. (2016). Combining self-affirmation and implementation intentions
to reduce heavy episodic drinking in university students. Psychology of Addictive Behaviors,
30, 434–441. [Link]
Peterson, C., & Seligman, M. E. P. (2004). Character strengths and virtues: A classification and
handbook. New York, NY: Oxford University Press.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and
comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–
891. [Link]
Scott, J. L., Brown, A. C., Phair, J. K., Westland, J. N., & Sch€
uz, B. (2013). Self-affirmation, intentions
and alcohol consumption in students: A randomized exploratory trial. Alcohol and Alcoholism,
48, 458–463. [Link]
 20448287, 2018, 1, Downloaded from [Link] by Kansas State University, Wiley Online Library on [10/03/2025]. See the Terms and Conditions ([Link] on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Combined alcohol intervention 127

Scott-Sheldon, L. A. J., Carey, K. B., Elliot, J. C., Garey, L., & Carey, M. P. (2014). Efficacy of alcohol
interventions for first-year college students: A meta-analytic review of randomized controlled
trials. Journal of Consulting and Clinical Psychology, 82, 177–188. [Link]
a0035192
Sniehotta, F., Presseau, J., & Ara
ujo-Soares, V. (2014). Time to retire the theory of planned behaviour.
Health Psychology Review, 8, 1–7. [Link]
Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of the self. Advances
in Experimental Social Psychology, 21, 261–302. [Link]
60229-4
Sugarman, D. E., & Carey, K. B. (2009). Drink less or drink slower: The effects of instruction on
alcohol consumption and drinking control strategy use. Psychology of Addictive Behaviors, 23,
577–585. [Link]
Voisin, D., Girandola, F., David, M., & Aim, M. A. (2016). Self-affirmation and an incongruent drinking
norm: Alcohol abuse prevention messages targeting young people. Self and Identity, 15, 262–
282. [Link]
Webb, T. L., Joseph, J., Yardley, L., & Michie, S. (2010). Using the internet to promote health
behavior change: A systematic review and meta-analysis of the impact of theoretical basis, use of
behavior change techniques, and mode of delivery on efficacy. Journal of Medical Internet
Research, 12, e4. [Link]
Wechsler, H., Lee, J. E., Kuo, M., Seibring, M., Nelson, T. F., & Lee, H. P. (2002). Trends in college
binge drinking during a period of increased prevention efforts: Findings from four Harvard
School of Public Health study surveys, 1993–2001. Journal of American College Health, 50,
203–217. [Link]
Wood, W., Tam, L., & Witt, M. G. (2005). Changing circumstances, disrupting habits. Journal of
Personality and Social Psychology, 88, 918–933. [Link]

Received 7 December 2016; revised version received 23 August 2017

Supporting Information
The following supporting information may be found in the online edition of the article:
Data S1. Moderation analyses.
Data S2. Intention-to-treat analyses.
Data S3. Negative binomial generalized linear mixed models.