Survey design


Here, we consider how the propositions of the NPT can be operationalised in terms of study designs that involve the use of quantitative research methods and focus on (i) phenomena of interest for study and (ii) developing and testing hypotheses.

As a middle range theory, NPT offers a framework for understanding the processes that are involved in the normalisation of new practices. As such, it represents some key propositions about the nature of the 'work' involved in normalisation - and in particular, about the activities that are involved in collective efforts to make new practices become normalised.

1. Phenomena of interest for study

The NPT's emphasis on normalisation 'work' that is collaborative has some important implications for the design and development of quantitative studies. The NPT recognises that individuals have an important role to play in contributing to activities that influence the potential for a new practice to become normalised. As such, individuals' perceptions (and the assessments that they are able to make with regard to a new practice) are important and worthy of data collection in studying normalisation from the perspective of NPT. This is in contrast to other theoretical approaches that focus more primarily on characteristics of the organisation in which changes of the kind that are of interest to the NPT take place. The first point concerning the phenomena of interest that is relevant to quantitative studies is that:

Individuals' own perceptions of a new practice are important and worthy of assessment

Although individuals' perceptions are, in themselves, important, the NPT is clear about what those perceptions relate to. The NPT takes as its starting point, the concept of 'the work'. Individuals' assessments are important in that they tell us something about the nature of the work involved in a new practice, from their perspective. For the design and development of quantitative studies that draw on the NPT it is important then to emphasise that the NPT does not focus on individuals' own intentions with respect to a practice that is of interest to the study. This is because the NPT does not see individuals' intentions and behaviours as the primary driver of normalisation - such phenomena is considered to be relevant to normalisation in a contributory sense but should not be focus of enquiry. Therefore, in collecting data from the perspective of NPT:

Assessment of individuals' perceptions should focus on the work involved in a new practice, rather than their own intentions or actions

The NPT acknowledges the collaborative nature of the work involved in normalisation processes. The kinds of normalisation activities that are of interest to the NPT generally involve diverse groups of individuals who have specific but contrasting roles with respect to a new practice. For example, a new healthcare technology might impact on the perceptions and behaviours of individuals with a range of perspectives, such as those of managers, doctors, nurses, technicians, administrators, and patients. For all kinds of 'participants', the normalisation of a practice involves particular kinds of work. However, to understand normalisation from the perspective of NPT:

Assessment of individuals' perceptions should include the perspectives of all groups of individuals who are affected by a new practice

For quantitative studies, this has several implications for the design and conduct of research studies:

  • Firstly, careful consideration must be given to sampling and recruitment to ensure that appropriate participants are included in the study. The development of appropriate sampling strategies for quantitative studies using the NPT thus requires gaining as much understanding as possible, about the kinds of participants involved, and their roles and/or potential relationships with the practice that is the subject of study.
  • Gaining insight into the roles of different groups of individuals is important not only for ensuring the best possible coverage of relevant perspectives in your study, but also for having sufficient knowledge to design relevant survey instruments (i.e. instruments that contain questions that are relevant to the work that people do).
  • In designing a quantitative study using the NPT it is possible that you will need to use different (or adapted) versions of sets of NPT derived questions as appropriate to different groups of individuals involved in the practice of interest. For example, some questions about Contextual Integration (component, within construct: Collective Action) that you might include in your survey as appropriate for individuals working in administrative roles are unlikely to seem relevant (or indeed may be unanswerable) for nurses, for whom issues about other aspects of Collective Action (components: Interactional workability; Relational Integration; Skill-set workability) are likely to be much more relevant in terms of their working practices. In developing the NoMAD instrument, we have taken these issues into account and offer further guidance on how to adapt the instrument for use in studies.

2. Developing and testing hypotheses

As a theoretical model, the NPT specifies a set of (empirically derived) constructs that represent mechanisms which are important in the normalisation of new practices, and a set of propositions relating to those mechanisms and how they contribute to normalisation. As such, NPT is amenable to the development and testing of hypotheses about the relationships between factors identified as influencing normalisation, and outcomes of normalisation efforts. Quantitative methods - utilising structured data collection from large samples of participants - is particularly well suited to hypothesis testing of this kind.

Things to consider

In designing studies for the testing of hypotheses based on the NPT there are some points that you may wish to consider:

  • The NPT proposes that the mechanisms of Coherence, Cognitive ParticipationCollective Action, and Reflexive Monitoring all contribute to normalisation in a facilitative sense. That is, the more positive the activity relating to each of these mechanisms, the more likely it is that the practice will normalise.
  • Thus, the NPT also does not specify that any particular mechanisms hold greater importance (or weight) than any others. The question of importance for different mechanisms remains open for empirical testing (though it is likely that such issues are highly context dependent).
  • The NPT also does not specify particular relationships between the mechanisms. It does not assume that the mechanisms operate in a linear or orderly fashion, rather it suggests that relationships between different mechanisms are likely to shift over time.
  • The NPT does not provide any particular definition of 'normalisation' for the purpose of measurement as an outcome variable for quantitative studies. The NoMAD instrument offers some example questions that may be used for this purpose. However, the concept of normalisation (as an outcome) depends very much on contextual factors relating to the practice itself, the environment in which it is operating, and the different groups of individuals that relate to it. In designing a research study to include measures of normalisation as an outcome for the purpose of hypothesis testing, it is necessary to develop your own measures based on what outcomes are relevant. These could be subjective (self-report) or objective (eg. Usage data). They are likely to be multiple. For example, just some normalisation 'outcomes' might be:
    • level of use
    • Increasing use over time
    • Amount of shift from one practice to another
    • Disappearance of a previous practice
    • Reported acceptability of a practice
    • Measures of quality of work stemming from use of the practice

Additional resources

For a presentation of a study that used NPT to develop questionnaire items, click here