Understanding Public Perceptions from Confidence Surveys involving Lancashire Constabulary

Abstract

Public perceptions are important in policing (Hohl et al., 2010; Jackson et al., 2012), but are
continuously elusive because only a few individuals come into direct contact with their local
police force (Bradford & Myhill, 2015). As such, surveys are often conducted in an attempt to
gauge public perceptions, but it is argued that this approach suffers from critical limitations
(Brown & Benedict, 2002).

The first issue is that there is no consensus over what the ‘dependent variable’ actually is, its
remit, interpretation by the public, and how questions should be worded. Consequently,
numerous surveys looking to gauge public perceptions often examine ‘confidence’,
‘satisfaction’, ‘trust’, ‘effectiveness’, ‘feelings of safety’, and ‘legitimacy’, without effectively
distinguishing the concepts beforehand (Cao, 2004; Luhmann, 1988). Secondly, there is also
no consensus over the measurement of variables, with varying use of Likert-type scales/options
for respondents, which may ultimately lead to form related errors (Albaum, 1997). The third
limitation relates to issues with chronology. As public perceptions are fluid, views are likely to
change over time. Because survey methodologies provide a ‘snapshot’ of a sample within a
target population, care must be taken in contextualising the findings to the timeframe from
which they were collected (Sindall et al., 2012). Finally, care is also needed in the geography
of findings. Results should be representative of the areas related to the respondents, as previous
research highlights how perceptions are largely governed by individuals’ immediate
environment (Williamson et al., 2006). Therefore, appropriate sampling and/or weighting
should be applied to ensure that the survey is generalisable and reliable (Dawson, 2016; Fotini
et al., 2013).

This report took account of the methodological issues highlighted in literature to contextualise
how findings from surveys involving Lancashire Constabulary could be interpreted for
practice.

It initially appeared that the surveys using national samples (which contained weighted
responses from force areas) reported higher agreement figures in their surveys in comparison
to direct samples of the Lancashire population. This formed a conflicting image of public
perceptions, which may have been indicative of sampling and weighing approaches. Whilst the
Lancashire only surveys sampled directly from the target population, there may have been
issues with non-response bias and an uncontrolled collection of more negative views.

Conversely, the national samples applied weighting to the participants based upon
demographics. Whilst this theoretically aims to make the views more representative of the
population, the lower number of respondents means that the estimates may not be fully robust
as positive views may have carried more weight purely on the basis of demographics.

However, a more likely reason for the difference in figures between the surveys was that the
questions could be interpreted as measuring different concepts. This related back to the most
critical limitation highlighted by Brown and Benedict (2002), who explain that there is no
consensus over what exactly is being measured. Based upon the questions from the surveys, it
is likely that three core concepts (confidence, satisfaction and effectiveness) may have been
interpreted by the public when responding.

Overall, from the five surveys that met the inclusion criteria, it is recommended that the Her
Majesty’s Inspectorate of the Constabulary and Fire and Rescue Services (HMICFRS)
commissioned survey (conducted by BMG research) should be cautiously interpreted to gauge
public satisfaction in 2018/19 (Dai & Johnson, 2009; Zhao et al., 2014). The Officer of
National Statistics (ONS) estimates of the Crime Survey of England and Wales (CSEW) should
be cautiously interpreted to represent public confidence in 2018/19 (Cao, 2015; Luhmann,
1988; Hart; 1988). In addition, both of these findings may be supported by the University of
Central Lancashire (UCLan) survey conducted in 2015. However, caution should be advised
in the use of this survey as it was not generalisable to full population of Lancashire and may
now be outdated. The Living in Lancashire (LiL) Survey and the report by the Office of the
Police Crime Commissioner (OPCC) were interpreted as measuring effectiveness. In this
instance, precedent should be placed upon the Living in Lancashire survey strand to cautiously
gauge local agency effectiveness (Ludwig et al., 2017), due to its methodological rigour.
Consequently, the OPCC survey should not be considered as providing a valid evidence-base
and should be used for information purposes only. This is due to limitations in the methodology
and analysis of survey data.