(2001) showed the theory of planned behaviour (TPB) to predict attendance at
individual screening appointments, but not to predict repeated attendance. In contrast,
Conner et al.(2002) reported the TPB to be predictive of long-term healthy eating
over a period of six years.
Recent theories have focused on the factors that might be different for initiation
and maintenance of health behaviours. For example, satisfaction with the outcome of
the behaviour (e.g. quitting smoking) may be important in the decision to maintain,
but not initiate a behaviour (Rothman, 2000). In contrast, self-efficacy may be an
important determinant of both initiation and maintenance behaviours but may act in
different ways (Bandura, 2000). For example, while high expectations may facilitate
initiation, these expectations must become more realistic in order that dissatisfaction
with outcomes during repeated performance does not inhibit maintenance (King,
Rothman and Jeffery, 2002).
There is a need to identify which change techniques are likely to enhance the
effectiveness of behaviour change interventions targeting particular behaviours among
specified groups (see Chapters 8 and 9). The meta-analytic approach utilized by
Albarracín et al.(2005) combined data from various studies of condom use inter-
ventions, which set new standards in exploring technique effectiveness and the extent
effectiveness was moderated by audience (e.g. young versus older people). Future
research will need to extend use of meta-analysis to other behaviours and to extend
the range of change techniques considered (see Abraham and Michie, 2008; Webb and
Sheeran, 2006). This includes use of meta-analyses to synthesize effect sizes observed
in experimental studies targeting particular change processes or determinants (Sheeran,
2012). Future trial design could also involve simultaneous tests of interventions with
and without particular technique combinations to establish whether the addition of these
techniques enhances effectiveness. For example, would the effectiveness of interventions
targeting attitudes towards increased exercise (e.g. by focusing on the consequences of
exercise behaviours) be enhanced by the addition of incentives and behavioural
contracts? By testing two or more versions of interventions (with and without the
additional techniques) such questions can be answered experimentally.
Another important area is the development and testing of behaviour change
techniques that are cost-effective when used with large populations. Although a large
number of change techniques have been identified some of these techniques can be
expensive to implement and may be challenging to sustain in practice because of the
range of competencies required to deliver them (e.g. motivational interviewing). Other
techniques may be simpler, easier to roll out to large segments of the population and
so cost-effective, even if they are associated with only small to medium effect sizes on
behaviour. We have considered a range of such change techniques including, for
example, self-monitoring, implementation intentions, changes in choice architecture,
regulation to introduce default opt-in systems and if-then planning techniques (see
Chapter 9). There is growing evidence that if-then planning can be used in large-scale
studies to encourage behaviour patterns such as blood donation (Godin et al., 2014)
and ongoing studies to suggest they might also be cost-effective in reducing smoking
intiation (Conner and Higgins, 2010; Conner et al., 2013a).
Similarly, a recent meta-analysis (Harkin et al., 2016) suggests that simple inter -
ventions can prompt large changes in self-monitoring (d+= 1.98) that are associated
FUTURE DIRECTIONS 259