The pesky problem of defining a ‘pest’: testing the pest management attitudes scale in the United Kingdom

ABSTRACT

The Pest Management Attitudes (PMA) scale was developed to provide a unidimensional and versatile tool to assess attitudes toward introduced pests and their management. While the PMA has been tested and shown strong psychometric properties in samples from Aotearoa, New Zealand (NZ), it is only beginning to be used internationally. This study tested the utility and influence of wording of the PMA scale in the United Kingdom (UK), using a 2021 survey (N = 999) distributed via online platform Prolific. Two of the 9 PMA scale items were not appropriate in our UK sample. We posit that despite references to introduced and native species in the PMA wording, many participants completed the survey with human rather than biodiversity pests in mind. While the PMA remains a valuable tool for understanding attitudes toward pests and their management, wording may need modification to ensure that concepts translate cross-culturally to retain meaningful comparisons.

KEYWORDS:

• Cross-cultural
• invasive species
• psychometric scale
• social dimensions
• wildlife attitudes

Introduction

Non-native, invasive species are often considered pests due to the threats they pose to biodiversity and human values toward biodiversity (Bellard et al., 2016). Management of such species is thus often undertaken. Yet this can result in social conflict, particularly where people’s attitudes are not well understood, or are not perceived to have been considered (Crowley et al., 2017; Redpath et al., 2013). Previous work on this issue has often focused on attitudes toward particular methods or species and has taken a context-specific approach (Bremner & Park, 2007; Courchamp et al., 2017; Fitzgerald et al., 2007; Kannemeyer, 2017; Russell, 2014).

Overcoming the prevalent context-specific approach, the Pest Management Attitudes (PMA) scale was developed to provide a unidimensional and versatile tool that could be used across a range of geographical contexts (Aley et al., 2020). The scale was initially tested using two studies in cities in Aotearoa, New Zealand (NZ). Psychometric validation, including exploratory and confirmatory factor analysis, confirmed a nine-item, one-factor model that worked in both studies (Study 1, N = 1190; Study 2, N = 1444). The mean PMA score was above the mid-point of three, indicating a tendency toward support for pest management, which was expected. Further validity tests showed that PMA score correlated in predictable ways with other measures of support for conservation and care for the environment, such as participation in conservation actions and the New Ecological Paradigm score (Dunlap et al., 2000). Given this success, the PMA scale has since been used in other NZ contexts (Kaine et al., 2021).

The PMA scale is beginning to be applied outside of NZ (Cerri et al., 2024), but for international use some of the PMA items may need to be adapted either because of language translation or context-of-meaning across cultures (e.g., the term “pest” versus other terms used in invasion science; Soto et al., 2024). Such changes should be piloted before deploying full surveys to ensure they maintain the intent and comparability of the PMA scale. The purpose of the present study was thus to assess the validity and results of the currently-worded English PMA scale when applied to a sample from the United Kingdom (UK), to provide further examination of the extent to which the scale has wider international utility as intended (Aley et al., 2020), and whether any wording modifications are required to enable international use. To maintain the most direct comparison with previous applications in NZ, prior to proposing any international wording changes, we used the same PMA question wording as used in previous NZ studies (Aley et al., 2020). We also included many other questions used in NZ studies, including the Environmental Concern Scale (Schultz, 2001) and a values questionnaire (Stern et al., 1998) to understand correlations between PMA with individuals’ environmental concern and value priorities.

The UK was selected because it is an English-speaking country, with cultural connections given NZ’s European colonization by British settlers in the 19^th Century. Despite these linguistic and cultural similarities, the countries are ecologically distinct. Whereas New Zealand is insular in character, with human-mediated species introductions only taking place over the past 700 years, the UK is continental in character, with human-mediated species introductions taking place since the Neolithic (Manchester & Bullock, 2000). Biodiversity in the UK has thus been altered by species introductions over a much longer period, with little evidence that those introductions have led to extinctions (Manchester & Bullock, 2000), with exceptions on offshore islands (Craik, 1997) and overseas territories (Hilton & Cuthbert, 2010).

Existing UK research suggests there may be less awareness of the role of non-native species in biodiversity loss, which translates to less support for their control or eradication (Bremner & Park, 2007; Philip & Macmillan, 2003). For example, a 2015 survey on attitudes toward introduced gray squirrels (which displace native red squirrels) revealed that 22% of respondents were unaware of the relationship between gray and red squirrels, and over half wanted to conserve both (Dunn et al., 2018). This contrasts with NZ, where awareness of non-native species and support for pest control are typically found to be high (Kannemeyer, 2017; Russell, 2014). Despite linguistic and cultural similarities, there is therefore potential for NZ and the UK to have distinct contexts in terms of pest management attitudes. The UK was thus a suitable first site to test the utility and influence of wording of the PMA scale outside NZ, while generating novel data UK residents’ attitudes toward pests and their management.

Methods

Participants

In November 2021, we distributed a survey to UK residents via the online crowdsourcing platform Prolific. An increasing number of studies use crowdsourcing platforms (Newman et al., 2021), which provide access to individuals who have signed up to complete online tasks for payment. Such platforms offer fast recruitment of large numbers of participants, for relatively low cost (Newman et al., 2021). However, issues have been raised around participant attentiveness (Hauser & Schwarz, 2016), fraudulent behavior, sample representativeness, data non-independence and in-group bias (e.g., participants may communicate about tasks) (Gray et al., 2016), and non-naivete of participants (Newman et al., 2021). Studies have typically shown that the data obtained by crowdsourcing is satisfactory, with some limitations and the need for quality-checking (Hauser & Schwarz, 2016; Newman et al., 2021).

Target sample size was 1000 so that a more or less statistically balanced comparison with previous NZ studies could be undertaken, with 999 valid responses received after the survey was closed having reached target. The completion time was at least two standard deviations below the mean of six minutes and 19 seconds (minimum completion time one minute and 22 seconds; maximum 42 minutes and 14 seconds), meeting Prolific’s criteria for inclusion. High attention was also indicated by all surveys being complete or near-complete, including the two qualitative text-entry open-ended questions for which only six individuals did not answer. We are therefore confident of reasonable data quality in our sample, despite known limitations of crowdsourcing platforms.

Survey

The survey was introduced by an information and agreement sheet, with participants asked to consent by ticking a box before proceeding. Participants were asked in which country they currently live, with only those selecting the UK permitted to proceed. All Prolific participants must be aged 18 or above. The survey was approved by the University of Auckland’s Human Participants Ethics Committee (024110) and adhered to the ethics of the American Psychological Association (APA, 2017).

We asked respondents demographic questions, and questions about conservation participation that were used in the original PMA study (Aley et al., 2020). While we recognize that there are cultural differences among nations and regions within the UK, these were not the focus of our study, so we did not collect regional information.

We used the identically worded final 9-item version of the PMA scale as reported in Table 7 of Aley et al. (2020). As in the original study, odd-numbered items were positively keyed, while even-numbered items were negatively keyed and reverse-coded for analysis. Participants indicated their level of agreement with each statement on a five-point Likert-type scale from strongly disagree to strongly agree.

Cross-cultural studies may require that terms be translated or changed to ensure understanding (Harkness, 2008). However, non-standardization, including of question wording, has previously undermined comparability of the internationally used New Ecological Paradigm (NEP) scale (Hawcroft & Milfont, 2010). Wording changes, even subtle, should therefore be tested to ensure that intent and comparability are maintained. We left PMA scale wording unchanged in order to, in the first instance, most directly test its suitability in the UK. We recognized that in the UK “invasive species” is more commonly used than “pest” to refer to introduced species that harm biodiversity (e.g., Forestry Commission, 2020). However, the PMA questionnaire itself stated that the statements were about “introduced pest species and their control,” and two PMA items mention native species (item three: “The benefits of pest control outweigh the risks to native species;” item five: “Native species have greater rights than pest species”). Nonetheless, anticipating that the term pest may be interpreted differently in the UK, we included two open-ended questions aimed at ascertaining which species and activities respondents had in mind during their response. These questions were presented at the end of the survey to avoid order effects that would undermine comparability.

Following the original NZ study (Aley et al., 2020), we included the Environmental Concern Scale developed to assess a tripartite model of concern-based environmental attitudes (Schultz, 2001). Respondents are asked, “I am concerned about environmental problems because of the consequences for…,” and indicate on a seven-point scale anchored at 1 (not important) and 7 (of supreme importance) their level of concern for 12 different factors, sorted into biospheric (animals, plants, marine life, birds), egoistic (me, my lifestyle, my health, my future), and altruistic (people in my country, all people, children, future generations) items. Items were presented in the same order for all participants. This scale has been used around the world and its tripartite model of concerns for environmental issues has been confirmed (e.g., Milfont et al., 2006).

A 15-item values questionnaire was included, based on that developed by Stern and colleagues (Stern et al., 1998). This questionnaire asks participants to rate the extent to which values act “as a guiding principle in my life,” on a scale from −1 (opposed to my values) to 7 (of supreme importance). Statements were aimed at measuring five value types, namely self-enhancement (authority, influential, wealth), conservatism (honoring parents, family security, self-discipline), openness to experience (a varied life, an exciting life, curious), altruistic (a world at peace, social justice, equality), and biospheric values (protecting the environment, unity with nature, respecting the earth). PMA and value scale items were presented in randomized order.

Statistical Analysis

Before conducting statistical analyses, we confirmed via the Kaiser-Meyer-Olkin statistic (.67) and Bartlett’s test of sphericity (p < .05) that UK PMA scale data were suitable for factor analysis. An exploratory factor analysis (EFA) was conducted using the default setting in Mplus, which uses maximum likelihood estimation and geomin rotation that allows factors to correlate. Mplus allows specification of the minimum and maximum number of factors to be extracted, and we examined 1 to 3 factor solutions.

Following initial EFA, we conducted further tests to determine which PMA items were most suitable for explaining results. These included Cronbach’s alpha to measure the association between items and thus assess the overall reliability of the complete test (Streiner, 2003) and repetitions of EFA on subsets of our data based on responses to qualitative questions. After determining which PMA items functioned well, we ran a confirmatory factor analysis in Mplus with maximum likelihood estimation with robust standard errors to confirm the validity of this model. A one sample, one-tailed t-test was used to assess whether average PMA score calculated using this model was significantly above the midpoint. Pearson’s correlations between PMA score and other survey questions, and environmental attitude and value scores (calculated by averaging across the relevant items in each scale) were calculated to assess differences in pest control support between sub-groups. Finally, to assess whether the model that was most suitable for explaining UK data would also have functioned well in NZ, we forced this model to a NZ sample from Study 1 of Aley et al. (2020).

Qualitative Analysis

Qualitative questions asked respondents: “Please list up to 5 species that you consider pests” (Q15) and “Please describe in your own words what ‘pest control’ means to you” (Q16). Analysis of species considered pests (Q15) involved normalizing terminology and extracting the most common pests, with each term excluding species within a category (e.g., “black rat,” “fire ant” were listed as separate terms) but including location information (e.g., “city mice,” “feral pigeons” (implying city-dwelling)). We assigned each pest both a taxon and nativeness status. Examination of some entries indicated misreading of the question as “pet,” not “pest.” Thus, for the purpose of analyses, we treated all mentions of horses, hamsters, snakes, and gerbils as not applicable, along with certain mentions of cats, dogs, rabbits, fish, and birds (i.e., when lists only contained companion species). In our examination of nativeness status, we listed rats, when named generically rather than a specific species, as their own category. We do this because while the most common rat, the brown rat (Rattus norvegicus), was introduced to the UK in the 1700s (Mammal Society, 2022b), we believe that this is not widely known; furthermore, the now uncommon black rat (Rattus rattus) has been present since Roman times (Mammal Society, 2022a). We included specific mentions of black or brown rats under non-native species. Similarly, we included rabbits when named generically as not applicable rather than non-native given their likely introduction by the Romans in the 1st Century AD (Mammal Society, 2022c). These examples convey the complexity of translating the PMA scale to a culture with a longer history of human-mediated species introductions.

Responses to personal meaning of pest control (Q16) were analyzed using inductive thematic coding. Each response was assigned up to nine codes based on its content in Excel. For consistency, coding was undertaken entirely by the first author. Steps taken to achieve intra-coder reliability included checking of the entire dataset for consistency after initial coding was completed, at which point codes referring to similar themes were consolidated. Those respondents who appeared to mistake “pest” for “pet” when naming pest species (Q15) were not excluded from analysis of personal meaning of pest control (Q16), as their answers indicated no misunderstanding.

Results

Respondents mostly described themselves as living in urban areas (63%), and as having lived in the UK for most of their lives (90%). After discarding invalid year of birth entries and participants who identified as gender diverse (who comprised too few in our sample, n = 8), we determined via Pearson’s chi-squared tests that our sample differed significantly, with medium to large effect sizes, to the UK’s 2020 mid-year population projections (Office for National Statistics, 2021) in terms of age (χ² (9, N = 988) = 471, p < .001, ϕ_c = .69) and gender (χ² (1, N = 990) = 172, p < .001, ϕ = .42). Examination of the data indicated that an over-representation of younger people, particularly those aged 30–34 (20% of the sample, compared with 9% in the UK population) and 35–39 (16% compared with 8%), and underrepresentation of people aged 65 and over (2% compared with 24%). The sample also had an overrepresentation of females (70.8%). These skews reflect typical biases of digital crowd-sourcing platforms and should be acknowledged when interpreting our results, although they may not ultimately affect the psychometric properties (Steelman et al., 2014).

Construct Validity

Average scores for each of the PMA items and fit measures are provided (Tables 1, 2). The three-factor solution was not easily interpretable and parallel analysis with 2000 simulated iterations indicated only two factors should be extracted (see Figure 1). While the two-factor solution had better model fit, it had significant cross-loadings for several items, and it did not represent any clear item grouping. The one-factor solution was thus deemed the more parsimonious model (Table 2). Results indicated that two items had weak or non-significant loadings: item five (.149, p < .05) and item six (.012, p > .05). Cronbach’s alpha values over 0.60 are typically considered adequate (Clark & Watson, 1995). The alpha value was 0.56, indicating only near-adequate association between items in the scale.

Figure 1. Examination of fit measures of the PMA scale from a parallel analysis with 2000 simulated iterations (n = 999).

Table 1. Descriptive statistics for the PMA Items, with ratings reverse-coded for even-numbered statements (n = 999).

PMA item	M	SD
1. Pest species are a significant conservation problem	3.41	0.82
2. Pest control interferes with nature	2.76	0.96
3. The benefits of pest control outweigh the risks to native species	3.14	0.82
4. Pest control has unknown side effects	2.66	0.80
5. Native species have greater rights than pest species	3.60	0.91
6. Today’s pest control methods are NOT proven to be effective	3.11	0.72
7. Investment in pest control is beneficial for future generations	3.68	0.73
8. Pest control is less important than other conservation issues	3.07	0.94
9. NOT enough pest control is being done	3.10	0.84

Table 2. Examination of fit measures from an exploratory factor analysis of the PMA scale items using maximum likelihood in Mplus (n = 999).

Factors	Eigenvalues	Chi‐Square Value	DF	Chi‐Square p-value	RMSEA	SRMR	CFI	Negative residual variance
1	2.23	302.798	27	<.0001	.101	.068	.70	no
2	1.43	124.01	19	<.0001	.074	.036	.89	no
3	1.08	38.06	12	.0002	.047	.018	.97	no

RMSEA = Root Mean Square Error of Approximation; SRMR = Standardized Root Mean Square Residual; CFI = Comparative Fit Index.

These findings, and inspection of responses to qualitative questions, led us to investigate the two non-significant loadings further. We observed that about a quarter of respondents mentioned a non-native species when naming pest species (Q15) or describing the personal meaning of pest control (Q16) (n = 243), compared with those who did not mention non-native species (n = 750) or did not provide an answer (n = 6). We repeated the previous analyses on two subsets of the data, distinguishing those who included non-native species in their qualitative responses (i.e., knowingly interpreted the PMA scale as applying to non-native species) from those who did not (i.e., conservatively assuming no mention implied potential mis-understanding of the context of the PMA scale). Notably, items five and six had stronger loadings when considering only those who mentioned non-native species (i.e., item 5: .228, p < .05; item 6: −.260, p < .05) than for the remaining participants (i.e., item 5: .110, p < .05; item 6: .035, p > .05). However, since items five and six did not function well for the overall UK sample, we focused on the seven-item scale for subsequent analyses. For the larger subset of those who did not mention non-native species in qualitative answers, this seven-item version of the PMA scale showed an overall better fit [χ²(14) = 101.615, p < .0001, RMSEA = .091, SRMR = .053, CFI = .85] than the one-factor model reported in Table 2.

A confirmatory factor analysis for the seven-item version of the PMA scale had reasonable model fit [χ² (14) = 82.33, p < .0001, RMSEA = .080 [.064; .098], SRMR = .053, CFI = .849, TLI = .77], but inspection of modification indices indicated residual covariance between items one and seven. Allowing residual covariance between these two items improved model fit: χ²(13) = 53.40, p < .0001, RMSEA = .064 [.047; .082], SRMR = .044, CFI = .91, and TLI = .86. Forcing this model to a NZ sample from Study 1 of Aley et al. (2020) yielded a poor model fit: χ²(13) = 229.23, p < .0001, RMSEA = .118 [.105; .132], SRMR = .061, CFI = .80, and TLI = .67. In this NZ sample, another item pair (8 and 9) had substantial residual covariance, and allowing only this covariance improved fit markedly: χ²(13) = 105.77, p < .0001, RMSEA = .077 [.064; .091], SRMR = .042, CFI = .91, and TLI = .86. The Cronbach’s alpha values for the seven-item version of PMA scale were acceptable in the UK (0.61) and NZ (0.67) samples. (Table 3 presents descriptive statistics of the other scales considered.)

Table 3. Descriptive statistics for the environmental concern and values scales (n = 999).

Scale	M	SD	Cronbach’s alpha
Biospheric concern	5.49	1.06	.89
Egoistic concern	4.98	1.18	.88
Altruistic concern	5.64	1.11	.90
Self-enhancement values	2.67	1.39	.70
Conservatism values	4.74	1.23	.60
Openness values	4.22	1.35	.75
Biospheric values	4.64	1.47	.86
Altruistic values	5.24	1.27	.71

Responses for the environmental concern scale ranged from 1 (not important) to 7 (of supreme importance) and those for the values scale ranged from −1 (opposed to my values) to 7 (of supreme importance).

Overall, these results indicated that the one-factor model of the PMA with the retained seven items performed comparatively well in both UK and NZ samples. However, two items performed differently for UK participants, particularly those who did not mention non-native species, and the associations between item pairs differ in the UK and NZ contexts.

Associations Between PMA and Other Scales

The average score for the seven-point PMA scale (M = 3.12, SD = .46) was above the midpoint of three, t(988) = 7.933, p < .001, with a small effect size (d = .25), indicating a tendency toward support for pest control in the UK, as was found in NZ (M = 3.63, SD = .61). Unlike in the NZ studies (Aley et al., 2020), there were no significant relationships between gender or age and PMA score. Few participants indicated that they are members of environmental or conservation organizations (4%) or have participated in conservation in the last 12 months (10%), and there was no correlation between PMA score and these responses (r = .303) (Table 4). A quarter (25%) of participants indicated willingness to be involved in pest control in their local area; there was a significant correlation, with a small effect size, between this and PMA score (Table 4). In both the environmental attitudes and general values scales, biospheric concerns and values were significantly negatively correlated with PMA score, with small effect sizes (r = −.148, −.184, respectively). In contrast, PMA score was positively associated with egoistic environmental values, as well as conservatism, openness, and self-enhancement, though effect sizes were small (r ≤ .213) (Table 4).

Table 4. Pearson’s correlation coefficients between seven-item PMA score and other variables.^a

Question	Question wording	r	p
Q6	Are you a member of an environmental or conservation organization? (1 = y, 2 = n)	−.030	.347
Q7	Have you actively participated in conservation in the last 12 months? (1 = y, 2 = n)	−.019	.547
Q8	Thinking generally, would you consider being directly involved in pest control for your local area? (1 = y, 2 = n)	−.125	<.001*
Q9	What gender do you identify? (1 = male, 2 = female)	−.053	.095
Q10	In what year were you born?	−.029	.361
Q12	Do you live in a more urban/city area, or a more rural/country area? (1 = urban/city, 2 = rural/country)	−.039	.215
Q13	As a guiding principle in my life, this value is … authority, influential, wealth (Self-enhancement)	.213	<.001*
Q13	As a guiding principle in my life, this value is … honoring parents, family security, self-discipline (Conservatism)	.181	<.001*
Q13	As a guiding principle in my life, this value is … a varied life, an exciting life, curious (Openness)	.085	.008*
Q13	As a guiding principle in my life, this value is … a world at peace, social justice, equality (Altruistic)	.009	.777
Q13	As a guiding principle in my life, this value is … protecting the environment, unity with nature, respecting the earth (Biospheric)	−.148	<.001*
Q14	I am concerned about environmental problems because of the consequences for … animals, plants, marine life, birds (Biospheric)	−.184	<.001*
Q14	I am concerned about environmental problems because of the consequences for … me, my lifestyle, my health, my future (Egoistic)	.126	<.001*
Q14	I am concerned about environmental problems because of the consequences for … people in my country, all people, children, future generations (Altruistic)	.047	.143
Q15 & Q16	Participants who mentioned non-native species in qualitative responses (0 = no mention, 1 = mention)	.054	.092

^a“Other” responses were excluded due to a low sample size. *p < .05.

Qualitative Results

Invertebrates (insects, worms, gastropods, or arachnids: 49%), rats (18%), and mice (10%) were the most commonly identified pests, with other mammals (10%), birds (6%), plants (3%), and other types of animals (4%) making up the remainder. The majority of pest species named were either native, or of unspecified origin (69%), while only 7% were identifiably non-native and present in the UK (excluding generic mentions of rats). A further 3% of identified pests are not found in the UK, and a further 3% of entries were deemed not applicable (including rabbits).

Table 5 lists the most common 15 codes featured in responses personal definitions of pest control (Q16), and the percentage of responses they featured in. Effects of pests on environments; other species; personal property; people in general; industries and agriculture; and health were all relatively common, with 9–14% of responses featuring these themes. Twelve percent of responses mentioned non-native species. Two percent (n = 24) of responses indicated some kind of disapproval of pest control, e.g., a sense that “pest” is a socially constructed and arbitrary category, that pest control is a lazy solution to a problem we caused, or that “pests” are innocent and do not deserve to be killed.

Table 5. Percentages of responses to the question (Q16): “Please describe in your own words what ‘pest control’ means to you,” featuring the most common 15 codes (n = 990).

Code	Percent (%)
Manage, control, or get rid of (in general)	23
Reducing or managing pest populations	23
Non-specific harmful/undesirable effects of pests	18
Kill or cull pests	15
Effects of pests on environment/ecosystem	14
Effects of pests on other species	14
Animal pests	14
Effects of pests on me, my home, or personal property	13
Relocate or remove pests	12
Invasive and non-native pests	12
Effects of pests on people in general	11
Reducing or preventing harm done by pests (impact type unspecified)	10
Pests overpopulating, overrunning, or infesting	9
Effects of pests on industries, agriculture, infrastructure, buildings	9
Effects of pests on health, disease, and safety	9

Discussion

Overall this study found that the PMA scale as developed and applied in NZ did not function equivalently in the UK. A key result was that two of the nine PMA scale items were not appropriate for this UK sample. This result has also been supported in a study using the PMA scale in Italy where these two questions were also excluded, although only assumptively (Cerri et al., 2024). Our results provide the evidence that indeed these two questions may not be appropriate when applying the PMA scale in Europe. The lack of fit of item five (“Native species have greater rights than pest species”) may reflect the finding discussed below that “pest” was widely interpreted to refer to human nuisance pests, despite wording in the survey to indicate a focus on non-native (introduced) pests. Thus, responses to other questions about pest management likely did not relate to respondents’ views about the relative rights of native and non-native species. Item six (“Today’s pest control methods are NOT proven to be effective”) also did not align well with the other scale items. There is no strong reason to expect that views on method efficacy necessarily to align with statements about values, although in NZ item six aligned well with the other scale items (Aley et al., 2020). The contrast with NZ could again relate to what kinds of “pests” respondents had in mind, and the kinds of pest control tools used in the UK compared with NZ (e.g., commensal pest control in the UK, controversial aerial toxins in NZ) (Kannemeyer, 2017; Russell, 2014).

Associations also differed between the sub-sample of respondents who mentioned non-native species in their qualitative responses compared with those who did not, with items five and six having stronger loadings in the sub-sample that mentioned non-native species. Furthermore, in the overall sample, there was a negative correlation between acceptance of pest control and care for the environment. Rather, PMA score was positively correlated with egoistic attitudes to the environment, as well as various values such as conservatism. This contrasts with NZ results, where PMA score was positively associated with other measures of support for conservation. For the sub-sample mentioning non-native species, the positive correlation between PMA score and egoistic values held, but the negative relationships between PMA score and biospheric environmental concerns were no longer significant, lending support to the interpretation than respondents had commensal rather than biodiversity pests in mind. The apparent reversal of the relationship between conservation concern and support for pest control compared with NZ may also have been affected by our sample bias toward younger and female participants in the UK, who in other studies have been least likely to support lethal pest control (Akiba et al., 2012; Bremner & Park, 2007; Dunn et al., 2018; Kannemeyer, 2017). However, neither gender nor age was significantly associated with PMA score.

Overall, we suspect that the above results may reflect a widespread interpretation in the UK of the term “pest” as referring primarily to human commensal pests rather than introduced pests as intended and specified in the original PMA score wording. This hypothesis is further supported by qualitative results, with only around a quarter of respondents mentioning non-native species (excluding generic mentions of rats). Rather, species most commonly mentioned were similar to those animals typically considered pests in UK homes and gardens as described by Baker et al. (2020) dominated by invertebrates, rats, and mice. This issue highlights the need to ensure that terminology and concepts are understood by the culture under study (Harkness, 2008). Even among invasion scientists, these terms are still contested (Soto et al., Preprint).

We therefore conclude that the PMA scale requires wording modifications to ensure that its core concepts are understood even in English-speaking contexts. Further testing of different terms should be conducted in the UK and NZ, and other English-speaking countries to determine whether any terms generate similar results across countries. In addition, further explanation at the beginning of the survey could be valuable, potentially using examples from the context under study (e.g., gray squirrel, mink, Japanese knotweed in the UK), provided that comprehension-checking questions are included to ensure that respondents read the explanation. Alternatively, or in addition, the scale could always be used with accompanying concept-checking questions such as the qualitative questions we used. These modifications could still be insufficient in contexts where the concept of non-native species as problematic is extremely unfamiliar. To understand whether this is the case, we suggest that the PMA scale be further tested in both locations where this is expected to be a risk and in countries where non-native pests are expected to be widely perceived as biodiversity problems such as in Australia (Fitzgerald et al., 2007).

Acknowledgment

We wish to thank the funding bodies that supported this research: Predator Free 2050 Ltd, which supported AP’s research via a Capability Development postdoctoral award, and the Rutherford Discovery Fellowship, which supported JR.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Data Availability Statement

Data remain confidential, as this was a condition upon which informed consent was obtained. We invite direct contact with the researchers to discuss data sharing for replication purposes.

Additional information

Funding

This work was supported by Royal Society of New Zealand Te Apārangi under JR’s Rutherford Discovery Fellowship; and Predator Free 2050 Limited under AP’s Capability Development Funding (Predator Free 2050 Limited).