outcome evaluators is needed to optimize intervention development in pilot and
feasibility studies, while blinding and firewall procedures may be needed in definitive
tests of a finalized intervention.
TAXONOMIES OF INTERVENTION CONTENT CHARACTERISTICS
AND EVIDENCE OF EFFECTIVENESS FROM DATA SYNTHESES
We have noted that our behaviour is regulated by processes operating at different levels,
from the intra-personal to international (see Figure 9.2) and that many distinct
regulatory processes operate within these levels, including, at the intra-personal level,
both reflective and impulsive processes. In stage 3 of the intervention mapping
process, designers identify techniques that will effectively alter these regulatory
processes. Researchers have developed useful lists, or taxonomies, of change tech-
niques. Abraham and Michie (2008) named and defined 22 frequently employed
change techniques linked to various theoretical accounts of regulatory processes
as well as four packages of change techniques used together in more complex
interventions. The utility of the taxonomy in categorizing intervention content was
tested using 195 descriptions of interventions provided in scientific papers identified
by three separate reviews, as well as a sample of 13 pairs of descriptions taken from
published papers and detailed manuals. Both the taxonomy developers and trained
coders were able to use technique definitions reliably to identify techniques in
intervention descriptions. This taxonomy has been widely applied (see for example,
Michie et al., 2013) and further work has confirmed that it defines techniques that
occur frequently in behaviour change interventions (Abraham et al., 2015). The
taxonomy includes four techniques derived from control theory and goal theory
(Carver and Scheier, 1982; Locke and Latham, 2002), that is, goal setting or intention
formation, self-monitoring, providing feedback on performance (that is, comparing it
to a set standard or goal) and review and revision of previously set goals, to facilitate
further goal setting. It also includes techniques designed to change the cognitive
determinants of motivation including providing information about the consequences
of action and about others’ approval of the action (derived from the theory of planned
behaviour and other similar models – see Chapter 7), as well as techniques useful for
learning new skills such as providing instructions and modelling (derived from social
cognitive theory). Four techniques based on operant conditioning processes were also
included, that is, providing rewards directly following performance of the target
behaviour (i.e. contingent rewards or reinforcement), teaching people to use
environmental cues to prompt change (e.g. placing medication by one’s toothbrush),
setting up a behavioural contract that the individual signs to commit themselves to
undertaking specified behaviours over a period of time and prompting practice, that
is repeating a newly learned behaviour so that it becomes well learnt and habitual (and
can be automatically triggered with little reflective processing). The taxonomy also
includes four commonly employed technique packages, namely, relapse prevention
(Marlatt and Donovan, 2005), stress management, motivational interviewing (Rollnick
and Miller, 1995) and time management. Longer lists of change techniques have been
provided by a series of subsequent taxonomies, some tailored to specific behaviour-
change problems such as promoting sexual health (Abraham et al., 2012) or promoting
CHANGING BEHAVIOUR 219