Reporting and Justifying the Number of Interview Participants in Organization and Workplace Research
| Date | 01 October 2016 |
| Author | Mark N. K. Saunders,Keith Townsend |
| Published date | 01 October 2016 |
| DOI | http://doi.org/10.1111/1467-8551.12182 |
British Journal of Management, Vol. 27, 836–852 (2016)
DOI: 10.1111/1467-8551.12182
Reporting and Justifying the Number
of Interview Participants in Organization
and Workplace Research
Mark N. K. Saunders and Keith Townsend1
Birmingham Business School, University of Birmingham, 116 Edgbaston Park Road, Edgbaston,
Birmingham B15 2TY, UK, and 1Grith Business School, Grith University, Nathan Campus, Kessells Road,
Nathan 4111, Australia
Emails: m.n.k.saunders@bham.ac.uk, k.townsend@grith.edu.au
In this paper we examine established practice regarding the reporting, justification and
number of interview participants chosen within organization and workplace studies. For
such qualitative research there is a paucity of discussion across the social sciences, the
topic receiving far less attention than its centrality warrants. We analysed 798 articles
published in 2003 and 2013 in ten top and second tier academic journals, identifying 248
studies using at least one type of qualitative interview. Participant numbers were con-
tingent on characteristics of the population from which they were chosen and approach
to analysis, but not the journal, its tier, editorial base or publication year, the interview
type or its duration. Despite lack of transparency in reporting (23.4% of studies did not
state participant numbers) we reveala median of 32.5 participants, numbers ranging from
one to 330, and no justification for participant numbers in over half of studies. We dis-
cuss implications and, recognizing that dierent philosophical commitments are likelyto
imply diering norms, oer recommendations regarding reporting,justification and num-
ber of participants. Acknowledging exceptions, dependent upon study purpose and data
saliency, these include an organization and workplace research norm of 15−60 partici-
pants, alongside credible numbers for planning interview research.
Introduction
Within organization and workplace (O&W)
empirical studies, qualitative interviews are a
central technique, being employed frequently
(Crouch and McKenzie, 2006) as ‘reliable gate-
ways’ into researching organizations (Alvesson
and Ashcraft, 2012, p. 240). Invariably, the utility
of such qualitative research interviews depends
on the participant or participants chosen in terms
of their coverage and the quality of data within
their responses (Alvesson and Ashcraft, 2012). In
planning and operationalizing research, sucient
participants need to be identified and chosen to
The researchers thank Dr Sue Ressia(Grith University)
for her assistance in collecting data for thispaper.
provide the breadth, depth and saliency of data
necessary for authentic analysis and reporting
(Curtis et al., 2000; Lincoln and Guba, 1985) and
to enable new insights and rich understandings
(Patton, 2015). Yet, a recent review paper for
the UK’s National Centre for Research Methods
(Baker and Edwards, 2012) highlights a lack
of explicit discussion across the social sciences
regarding how many qualitative interview partici-
pants are deemed sucient and, along with others
(Curtis et al., 2000; Robinson, 2014), we suggest
that this topic has received far less attention than
its centrality warrants.
What is considered methodologically valid (au-
thentic and credible) diers between communi-
ties of qualitative scholars (Baker and Edwards,
2012) with diverging philosophical commitments
© 2016 British Academy of Management. Published by John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4
2DQ, UK and 350 Main Street, Malden, MA, 02148, USA.
Reporting and Justifying Interview Participant Numbers 837
(Johnson et al., 2006), research being subject to
peer review, (consensual) judgement and valida-
tion (Patton, 2015) by epistemological gatekeepers
(Symon and Cassell, 1999). Acknowledging this,
we commence with an overview of the literature
regarding the number of interview participants
likely to be sucient, reviewing both empirically
grounded evidence and expert opinions. This re-
veals a paucity of empirical research, highlighting
the need to establish accepted practices regarding
the reporting, justification and number considered
sucient within the broad network of communi-
ties of those undertaking O&W research. Within
our broadly pragmatistphilosophy we believe such
knowledge can be useful in informing the prac-
tice of qualitative researchers, whatever their epis-
temological and ontological positions, providing
insights into current practice and enabling recom-
mendations to support the planning, operational-
ization and reporting of future research.
Using 248 O&W empirical studies in 244 articles
drawn from a sample of ten top and second tier
research journals we examine established practice
judged worthy of publication. In adopting a
neo-empirical inductive position we engage with
the data provided in these articles as empirical
evidence analysing it systematically to draw
out recommendations regarding the reporting,
justification and number of qualitative interview
participants. Within our analysis we consider pos-
sible dierences in reporting practice contingent
on the journal, and its tier, editorial base and year
of publication, the type of qualitative interview
and interview duration, the populationfrom which
participants were chosen and approach to analy-
sis. Finally based on the practices reviewed, and
recognizing that diering philosophical commit-
ments will be reflected in research practice,we oer
recommendations regarding the reporting, justifi-
cation and number of participants. These include
both a broad range for the number of qualitative
interview participants and estimates of participant
numbers for planning interview research that are
likely to appear credible within O&W research.
Qualitative research interviewing and
the number of participants
Reporting practices for qualitative interviews
Qualitative interviews are used in a variety of
research designs (Brinkmann and Kvale, 2015;
Crouch and McKenzie, 2006). Unlike interviewer-
administered questionnaires, they comprise a rel-
atively free-flowing interchange of views between
two or, in the case of group interviews, three
or more people (Saunders, Lewis and Thornhill,
2016). Such interchanges are diverse in nature,
varying between disciplinary roots and dependent
upon epistemological and ontological assump-
tions (Johnson et al., 2006), being characterized by
philosophical diversity (Guba and Lincoln, 1994)
and within management research methodologi-
cal pluralism (Easterby-Smith, Golden-Biddle and
Locke, 2008). They also vary depending upon the
research purpose, comprising both breadth and
scope (Bryman, 2012), diering in the extent to
which questioning is unstructured, and the con-
duct and duration of the interaction between inter-
viewer and participant(s) (Brinkmann and Kvale,
2015). In comparison to quantitative research,
qualitative interviews are argued to oer greater
ecological validity, providing rich insightful ac-
counts and the ability to help make sense of com-
plex organizationalrealities (Eby, Hurst and Butts,
2009).
In reporting research drawing on qualitative in-
terviews, researchers are expected to explain and
justify their data collection and analysis trans-
parently in relation to their purpose (Baker and
Edwards, 2012; Robinson, 2014), thereby allow-
ing users to judge its utility.Dominant publication
conventions, albeit neo-positivist (Alvesson and
Ashcraft, 2012), dictate a description of method,
outlining concisely how participants were chosen
and data collected, and providing appropriate re-
flexive acknowledgement regarding bias or con-
flicts of interest to aid transparency (Meyrick,
2006; Robinson, 2014). This implies a need to
state the number and characteristics of partici-
pants interviewed (Miles, Huberman and Salda˜
na,
2013; Patton, 2015) and reasons for their selection
(Marshall et al., 2013).
Literature regarding the number of participants
usually treats each interview as a discrete event
involving one or more participants. Participant
numbers are argued to depend on the balance be-
tween representativeness (in a loose sense) and
quality of responses (Alvesson and Ashcraft, 2012)
in obtaining sucient information. For some stud-
ies, such as where the purpose is to establish if
something is possible or to provide a rich ac-
count, a single qualitative interview (or case) is
argued to be appropriate (Becker, 2012; Patton,
© 2016 British Academy of Management.
To continue reading
Request your trial