Original Research Reports “You’re Important, Jeremy, but Not That Important”: Personalised Responses and Equivocation in Political Interviews

This study was an assessment of personalised equivocation in political interviews, namely, politicians’ responses to questions which, in lieu of an explicit reply, are directed personally at the interviewer. Twenty-six interviews with recent UK party leaders were analysed in terms of questions, replies, and personalisation. The majority of personalised responses contained elements of criticism, although over a quarter were more amicable. For the eight featured politicians, the use of such responses was adjudged to be more about individual communicative style than their position on the political spectrum. Only one politician did not respond in this manner, indicating a more widespread use of personalisation than was previously suggested. Furthermore, an evaluation of interviewer follow-ups showed its effectiveness as a diversionary tactic in the face of troublesome questions. In terms of the proportion of questions which receive a full reply, a general reply rate analysis highlighted how recent political leaders have changed little from their predecessors. of 2015 and 2017. The exchanges between interviewers and politicians were coded for question type and, for the politicians’ responses, whether a reply was forthcoming. Each response was assessed for personalisation, and our specific focus was those deemed equivocal – namely, those in which a full reply to the question was not provided. Seven percent of equivocal responses contained personalisation – although not all were critical in nature; over a quarter were more genial. Overall, most of the personal responses appeared to divert the interviewer from following up with a similar question. As for our reply rate analysis, altogether the politicians gave full replies to 38% of questions.

itate the quantitative analysis of interviews, Bull (1994) proposed a method to clarify identification of both questions and replies. Questions typically take an interrogative form (e.g., "Did you authorise that?"), but not always. Quirk, Greenbaum, Leech, and Svartvik (1985) refer to questions that are declarative in form, which are often voiced with rising intonation at the end of the statement (e.g., "You authorised that?"). Indeed, a request for information may not only be, syntactically, of non-interrogative form but even contain no verb. For example, in a 2017 interview of PM Theresa May, Jeremy Paxman was asking about net migration to the UK: Can you tell us what it is now? Paxman: Yes it was 248,000 in the last set of figures. May:

May:
Note. Source: Sky  Clearly, the second turn of Paxman here is a request for information, but contains no verb; these are referred to as moodless questions (Jucker, 1986). Harris (1991) suggested that political interviews include substantial numbers of questions that are non-interrogative in form, and that these are put forward by the interviewer for the politician to express agreement or disagreement.
Questions in interrogative syntax fit one of three distinguishable forms based on the type of expected reply (Quirk et al., 1985). Those seeking affirmation or negation are defined as yes-no (or polar) questions (e.g., "Will there be tax increases?"). Those seeking a reply from an open range are identified as wh-questions; these feature an interrogative word, namely, how (the only one not to begin with wh), who, why, when, which, or what (e.g., "What does that mean?") [Bull (1994) added where to this list]. Those which present two or more alternatives are referred to as alternative (or disjunctive) questions (e.g., "Was that the fault of the Chancellor, or are you to blame?").
In addition to the three interrogative types (yes-no, alternative, and wh-questions) and the two non-interrogative (declarative and moodless), Bull (1994) added a third non-interrogative type: indirect. Bull (2009, p. 217) defines indirect questions as "a means of asking questions through reporting that of another". He provided the following example (in which the question's force is not expressed directly, but is expressed via a subordinate clause): "Many people have asked the question why did you go to war in Iraq". Bull's (1994) analysis identified over 1,000 questions from 33 political interviews, and all were identifiable as one of these six question types.

Previous Research -Replies and Evasion
As questions are identifiable as a request for information, a reply is a response which provides the requested information. In terms of political interviews, assessment of replies is interesting in that it addresses the unflattering view which prevails in the public perception: that politicians, in general, are evasive (Harris, 1991). Harris's study was conducted to evaluate the extent to which politicians do give straight answers to questions. From 17 political interviews conducted between 1984 and 1987, mostly with leading UK politicians, Harris found that direct answers were provided to just over 39% of the questions asked. Bull's (1994) clarification of questions and replies in political interviews proposed three types of response to interviewers' questions: Replies provide, explicitly, the information requested; Non-replies are a failure to provide the information; Intermediate replies sit midway on a continuum between replies and non-replies. Intermediate replies include instances where the politician's answer is Journal of Social and Political Psychology 2020, Vol. 8(2), 560-581 https://doi.org/10.5964/jspp.v8i2.1095 not stated explicitly but is implied, instances where the politician gives only a partial answer, and instances where an assessment of whether a reply was forthcoming cannot be made because the politician was interrupted. Bull's analysis of 33 interviews with leading UK politicians conducted between 1987 and 1992 found that 45.6% of questions received an explicit reply. Noticeably, this figure, termed reply rate, is comparable to the figure of 39.3% which Harris (1991) reported for direct answers from an entirely different set of interviews. More up-to-date studies, although of only a small set of interviews (Bull, 2016(Bull, , 2017, are indicative of even lower reply rates by recent party leaders. All of the above studies are consistent in showing that leading UK politicians tend not to give full replies to the majority of interviewers' questions. Of course, studies of equivocation in interviews and press conferences extend beyond UK politics. Researchers have examined the practices of politicians worldwide, including, the United States (e.g., Clayman, 2001;Clementson & Eveland, 2016), Malaysia (Nur Zahraa & Siti Rohana, 2016), Montenegro (Vuković, 2013), Taiwan (Huang, 2009), and Turkey (Çakir, Kökpinar Kaya, & Kara, 2016). Such studies, however, either did not set out to evaluate reply rates or, where statistics are reported, they are not always suitable for comparison with the foregoing UK studies. A clearer comparison can be made with a study of Japanese politicians (Feldman, Kinoshita, & Bull, 2015), where, from a series of interviews conducted in 2012 and 2013, national politicians gave direct replies to 41.4% of questions. This figure is noticeably similar to the reply rates reported for UK politicians in the 1980s and 1990s (Bull, 1994;Harris, 1991).

Previous Research -Personalisation
The study of personalisation in politics has meant different things to different researchers. For example, Thornborrow and Montgomery (2010) utilise the term to relate to subjective experiences of interviewees. Leone (2013, p. 133) defines it as "the phenomenon of leaders' profiles eclipsing those of their respective parties". Here, the term is used for discourse directed at a person present, with the intention of having personal relevance to that individual.
This form of personalisation in political communication has been investigated in areas beyond the broadcast interview. For example, in Prime Minister's Questions (PMQs) -the UK Parliament's highest profile event -Waddle, Bull, and Böhnke (2019) investigated personalisation in the question-response exchanges between the leaders.
Their evaluation of what are, effectively, personal attacks showed that, of the five PMs between 1979 and 2016, David Cameron was the most personally aggressive. In one period of Cameron's premiership (the ten PMQs sessions before the 2015 General Election), over 60% of his responses to Leader of the Opposition (LO) Ed Miliband contained a personal attack. Whilst acknowledging the differences between the two modes of political communication, particularly the rivalrous situation which exists between the leaders, Waddle et al. proposed that, similar to political interviews, a potential function of personalisation in PMQs is equivocation.
An extensive literature search shows that evaluations of this form of personalisation have not been the primary focus of political interview research. However, close scrutiny of research into evasion does yield some findings. These tend to be from studies utilising the aforementioned typology of equivocation (Bull, 2003;Bull & Mayer 1993 resorted to responses which were highly personal. Simon-Vandenbergen suggested that, beyond challenging an interviewer's neutrality, responding with such personalisation was characteristic of extremist discourse. Subsequent research, conducted by Mehdipour and Nabifar (2011), included an analysis of evasive techniques over a series of 20 interviews held between 2001 and 2010. They used the typology of equivocation discussed above (Bull, 2003;Bull & Mayer, 1993) to evaluate the responses of two high profile figures in US politics: Condoleezza Rice and Hillary Clinton. They reported a total of 362 questions asked, 146 of which received an equivocal response. Only one of these equivocal responses -given by Rice -was identified as an attack on the interviewer.

The Current Study
The focus herein was an analysis of interviews with UK party leaders from two recent general election campaigns.
It is apparent from the foregoing reviews, that personalised equivocation is more diverse than the critical comments used by Thatcher, as reported by Bull and Mayer (1993). A detailed analysis of the responses of recent politicians would provide data to address a number of key areas: the range of personalisation in mainstream political discourse, the extent to which it forms an equivocal response (including measuring its effectiveness as an evasive tactic), and the recent party leaders who are inclined to use it. This study is not only an investigation into this specific form of dialogue, but also an overdue up-to-date assessment of how leading politicians respond to interviewers' questions.
ii Analysis included an assessment of interviewers' questions and politicians' responses in accordance with the methods of Bull (1994 categories of personalisation proposed by Waddle and Bull (2016). The use of that categorisation system -with its seven types of personalised replies -means that the scope of personalisation analysis is now broader than the single attacks the interviewer category characterised in responses by Thatcher in the 1980s. Therefore, it was predicted that personalised equivocation levels would be higher in these recent interviews. For the same reason, there was an expectation that more politicians would be observed using personalisation as an equivocal response than had previously. It was also predicted, as suggested by recent studies of single interviews (Bull, 2016(Bull, , 2017, that the overall reply rate of current politicians would be lower than those of their predecessors.

Interview Participants
The politicians who featured in the analysed interviews were all leaders of UK political parties at that time.  Table 1.

477:00 Total
Note. Interviews by the BBC's Andrew Marr with the leaders of UKIP were shorter in duration than the others, but, during both campaigns, they were interviewed twice. Interviews by Jeremy Paxman -broadcast as The Battle for Number 10 on Sky News and Channel 4 (C4)were conducted on both occasions with only the two leaders who might be elected PM. Duration = minutes:seconds.

Apparatus
Videos of the interviews were accessible from either bbc.co.uk or youtube.com. Online searches (detailed below) led to the following websites for interview transcripts: bbc.co.uk, facebook.com, news-watch.co.uk, politicshome.com, skygroup.sky, and spectator.co.uk. Online transcripts could not be located for three interviews; these were transcribed by the first author. [See Appendix for transcript web page details.]

Procedure
General election campaigns were selected as the basis of the analysed interviews because, at such times, broadcasters typically include all major party leaders. Also, particularly for leaders of the largest parties, airtime tends to be similar. General elections in the UK can be as much as five years apart, but the two most recent these were not available, they were transcribed by the first author. Using the videos and the transcripts, the interviews were analysed for personalisation and reply rate separately. Both analyses (detailed below) were conducted in full by the first author. In each case, initial analysis was based on the transcripts. To facilitate accuracy, anything even vaguely ambiguous from the transcript alone was checked using the videos.

Personalisation
Personalisation analysis involved assessing each of the politicians' responses for comments directed personally at the interviewer. [Similar analyses have also been conducted by the first author on exchanges in PMQs, which an interrater reliability test (Cohen, 1960) showed to be highly reliable (k = .88, p < .001) ].
Instances of personalisation were then coded in accordance with the categories identified by Waddle and Bull (2016), as listed in Table 2. Seven categories of personalisation were identified in that study; each one is defined below.
Journal of Social and Political Psychology 2020, Vol. 8 (2)  Five of the categories are distinctly critical: effectively, personal attacks. The first of these -interviewer biasrelates to suggestions that the interviewer has, or is displaying, biased personal or political views. An example of this can be seen in a response to a question by Labour politician Dennis Skinner. Skinner was being questioned by the BBC's David Dimbleby in 1992 about Labour Party policy on the voting system proportional representation (PR). Skinner's reply included an accusation of bias towards PR: "Well you ought to have explained it much more clearly then Mr Dimbleby. I know you're one of the chattering classes that believes in proportional representation".
The category broadcast organisation bias relates to instances where personally directed comments are critical of the organisation represented by the interviewer, particularly in relation to programme content and coverage.
Dennis Skinner followed up the above with a response identified as such: and we had it rammed down our throats. We had it rammed down our throats for about three week [sic].

Every time Paddy Backdown v opened his mouth about it, you went on about it and turned it into an agenda
Skinner: item. The truth is that Ashdown and his party got eighteen percent. Proportional representation has not been supported in this election. And I'm fed up of hearing on your programme today, all these people trotting out that, that the Labour party's got to get in bed with the Liberals. The whole thing is a nonsense. […] Interviewer competence is the category which includes personal comments which are disparaging about the interviewer's intellect, knowledge, or their level of skill. The opening sentence voiced by Skinner in the first example

Reply Rate
For this analysis, it was first necessary to identify the interviewers' questions in accordance with the procedures for political interview research by Bull (1994), as described above. Thereby, each question was coded as either a yes/no, wh-, alternative, indirect, declarative, or moodless question. [In accordance with previous research (Feldman, 2016), small talk questions vi were not included.] Repeated or rephrased questions were counted as fresh instances. Some questions, though relatively rare, could be coded as more than one type (e.g., when the interviewer posed a double-barrelled question). Coding the questions in this manner facilitates an assessment of whether the requested information has been provided in the politician's responses. Thus, an evaluation was made as to whether the politician had made a reply, an intermediate reply, or a non-reply (Bull, 1994), also described above. This analysis was conducted in full by the first author. To confirm reliability, an interrater evaluation was conducted on over 15% of the corpus, which was also analysed independently by the second author. In terms of identification of questions, an interrater reliability test (Cohen, 1960) was applied and resulted in k = .83, p < .001.
Journal of Social and Political Psychology 2020, Vol. 8(2), 560-581 https://doi.org/10.5964/jspp.v8i2.1095 As for the identification of replies, a further test resulted in k = .77, p < .001. According to statisticians Landis and Koch (1977), the first of these represents "almost perfect" agreement, and the latter is very high on their range of "substantial" agreement (p. 165), confirming the high reliability of this analysis overall.

Results
The results of the reply rate analysis are shown in  The personalisation analysis was cross-referenced with the reply-rate analysis to evaluate the use of personalised politician directed at the interviewer within an explicit reply was thereby disregarded. For example, in the following extract, Jeremy Paxman was pressing Ed Miliband on whether he would make deals with the Scottish National Party. Miliband bantered with Paxman, but as he also made a direct reply here (answering "No" to a yes-no question), this personalisation was not accounted for.
You are, if you have any chance of forming a government you will won't you? Paxman: No, don't be so presumptuous. We've got six weeks to go, six weeks to go, you don't get to decide the election results six weeks before the general election. You're important Jeremy but not that important. It's the British people who decide.

Miliband:
Note. Source: Sky News & C4, 2015. Table 4 shows the results of the personalisation analysis. Overall, 7% of the 544 responses which were not full replies contained personal comments directed at the interviewer. The politician with the highest level of personalisation on this basis was Nigel Farage, with 17.39%. Percentage rates for other politicians were: Nick Clegg -12.12%; Tim Farron -10.53%; Ed Miliband -9.68%; David Cameron -5.41%; Theresa May -5.36%; Jeremy Corbyn -3.51%. None of Paul Nuttall's equivocal responses contained personalisation. Note. Shows numbers of politicians' responses to questions which were not full replies and which contained personalisation. Percentages in parentheses. IR = Interviewer; BO = Broadcast organisation; comp. = competence; FoM = frame of mind; Bland. = Blandishments (e.g., flattery, banter).
All instances of personalisation from these 26 interviews were identifiable in accordance with the categories identified by Waddle and Bull (2016). The most common type was befitting the category interviewer conduct, with 44.74% identified thus. One such example occurred as interviewer Robert Peston pressed Tim Comments identified as being critical of an interviewer's competence accounted for 23.68%. One such example followed a question by Evan Davis about UKIP income tax policy; Farage responded with: "Well you ought to do your research a bit better because you are wrong about the top rate of tax." The following exchange includes the only response identified as befitting the category interviewer bias. Here, There was one response by Farage which was critical of the broadcast organisation (the BBC). However, this followed a question by the BBC's Andrew Marr asking Farage about the future of the corporation and, importantly, contained no element of personal criticism, so was not coded as this form of personalisation. There were no personalised comments making reference to the interviewer's frame of mind.

Farage's response appears to question the neutrality of interviewer Davis [IR indicates where speech by the inter
The final assessment for personalisation was to gauge its potential for effectiveness. This was achieved by analysis of the interviewers' follow-up: whether or not, following the personalisation, there was a clear repeat or reformulation of the question. Of the 38 instances of personalised equivocation, only 14 were followed by the same question or a reformulation of it.

Discussion
This study was focused on recent televised political interviews in the UK. Included were interviews with leaders of UK parties: all held and first broadcast during the 2015 or 2017 general election campaigns. The primary focus, and the novel contribution of this research, was an assessment of personalisation by the politicians, namely, personal comments directed at the interviewer in lieu of a direct reply to their question. To facilitate the evaluation of personalisation used in this way, an analysis of questions and replies was also necessary. The second of these analyses would highlight whether or not a question had received a full reply, and where this was not the case, these responses featured in the personalisation assessment. A consequence of the second analysis was an upto-date and long overdue evaluation of reply rate -the extent to which politicians answer interviewers' questions -from a series of interviews with leading politicians. The results of both the personalisation and reply rate analyses are discussed below. of these responses included personalised comments aimed at the interviewer. Waddle and Bull (2016) identified seven types of personalisation used by politicians in interviews. Five of these were evident from the current corpus.
These findings show highly personalised responses to be, not only characteristic of extremist discourse (Simon-Vandenbergen, 2008), but an occasional tactic of mainstream politicians. The most common -almost 45% -were comments about the interviewers' conduct. Over a quarter were of the more genial variety -labelled blandishments. Just short of a quarter were comments making reference to the interviewers' competence. There was only one example of a politician levelling an accusation of personal bias at the interviewer. There was also only a single example of a personalised response fitting the category interviewer history.
Findings support Waddle and Bull's (2016) personalisation typology in that all personalised equivocal responses were identifiable as one of the seven types therein. Notably absent from these interviews were responses befitting the category interviewer frame of mind. This specific form of personalisation was a reported feature of responses by Labour politician Peter Mandelson. One such case is highlighted in the Introduction above. Another occurred during an interview with the BBC's Tim Willcox in 2010, when Mandelson repeated the phrase "calm down" four times. A further instance of this highly personal response was evident during a 2010 press conference, where Mandelson and other leading Labour politicians were being questioned by journalists, including Sky's Adam Boulton. On that occasion, Mandelson repeatedly told Boulton to calm down. The obvious conclusion from these observations is that this form of response is not widely used, but is an occasional tactic of Mandelson's, and one which he uses somewhat successfully to achieve greater control of the conversation.
For the individual politicians in this study, UKIP leader Nigel Farage directed the largest proportion of personal comments at the interviewers. Over 17% of his equivocal responses contained personalisation. Interestingly, Farage's successor as UKIP leader, Paul Nuttall, directed no personal comments at the interviewers during his equivocal responses. Comparing these figures to previous findings (Bull, 2003;Bull & Mayer, 1993), only Farage was a more frequent user of personalisation than former PM Margaret Thatcher, whose reported rate of personalised equivocation was 13%. It is important to note that previous research did not encompass as broad a range of personalisation as this study, chiefly, the inclusion of more genial tactics (labelled blandishments). However, only one of Farage's personalisations was identified thus, so, even without blandishments, the former UKIP leader's personalisation was still slightly higher than Thatcher's.
As only one of the political leaders in this study used personalised equivocation at a level above that of former PM Thatcher, this does not represent support for the prediction of higher levels by the recent politicians. However, all but one of the eight politicians here occasionally resorted to personal comments in lieu of a direct reply; previous research (Bull, 2003;Bull & Mayer, 1993) reported personalisation tactics by only one of three leaders analysed.
So, as expected, more politicians were observed using personalised equivocal responses. This may be indicative of an increased likelihood for politicians to be personal in their responses. However, as the current method is broader in scope and the videos and transcripts of the interviews from the 1980s and 1990s were not at our disposal, this could not be confirmed.
A closer look at one of the politicians here highlights a notable distinction. At just over 5% of his equivocal responses containing personal comments, former PM David Cameron was one of the lowest users of personalisation across these interviews. In contrast, an entirely different picture emerged from an analysis in a parliamentary setting Journal of Social and Political Psychology 2020, Vol. 8(2), 560-581 https://doi.org/10.5964/jspp.v8i2.1095 . When responding to questions from his political opponent, his personal aggression then was highest of the five PMs analysed. In one period of ten consecutive sessions of PMQs, almost 62% of his responses contained a personal attack on the LO. The salience of this sharp difference in one politician's behaviour across two distinct modes of political communication is perhaps more noteworthy with consideration of the timings.
The PMQs sessions assessed by Waddle et al., and Cameron's interviews analysed here, both took place between January and April 2015. Clearly, the then PM had a tendency to respond with personal antagonism towards a political opponent; the same tendency was not a feature of his behaviour towards political interviewers.

Functionality
In terms of the effectiveness of personalised equivocation; namely, whether it has the potential to divert the interviewer from what the politician may deem a troublesome question, analysis showed that on 63% of occasions, the question was not repeated or reformulated. This falls somewhat short of the 83% reported for Thatcher's interviews following her personal attacks on the interviewer (Bull & Mayer, 1993). However, Bull and Mayer contrasted the effectiveness of that technique with some of the more defensive means of equivocation used by Thatcher's political opponent, Labour leader Neil Kinnock. Those varied from only 25% effective, to entirely ineffective (i.e., the interviewer always followed with a repeat or reformulation of the question). Taking account of those previous findings, the 63% result from this study suggests that personalised equivocation -playing the man, not the ball -has the potential to be an effective strategy in diverting the interviewer from a troublesome line of questioning.
Cameron's clear disparity in personalisation across the two modes of political discourse suggests that personalisation in PMQs serves a different purpose than in interviews. Some of the potential functions cited in Waddle et al.'s (2019) PMQs research are not transferable. For example, personal criticism may serve to magnify cognitive differences between political opponents (Ilie, 2004). Indeed, particularly in the run up to a general election, party leaders may be motivated to make personal attacks in an attempt to reveal shortcomings in their opponents and to promote themselves. Attacks on an interviewer suggest a somewhat different motivation.  also discuss how personalisation in PMQs might function as a means of equivocation. Ilie (2004) proposes that the emotional force of a personal attack outweighs its rational force. Conceivably, if personalisation has an emotional effect on an opposition politician, an interviewer might be similarly affected. Arguably, an interviewer does not expect to become the focus of a politician's response. The interviewer's role is to set the agenda (Greatbatch, 1986). Occasions when they become the topic of conversation might serve to distract them from their line of questioning. Indeed, even personalisation which is amicable may engender a similar reaction. Atkinson (1984) expressed such a view from his observation of former PM Harold Wilson. Wilson's reported personalisation was merely to address the interviewer by his first name -a tactic which could "neutralize, albeit temporarily, the forceful interviewing style of Robin Day" (p. 174).
Our finding highlighting the potential effectiveness of personalisation as an interview tactic is a strong justification, in the interests of transparency, for research of this nature and the importance of having a reliable coding system to assess its use by those who represent or seek to represent us.

Reply Rate
Reply rate -defined as the percentage of responses which explicitly provide the information as requested in the interviewers' questions -was also assessed. For the 2015 interviews this was 43%, for 2017 it was 34%, giving an overall reply rate across all 26 interviews of almost 38%. Research on interviews from the 1980s and 1990s Journal of Social and Political Psychology 2020, Vol. 8(2), 560-581 https://doi.org/10.5964/jspp.v8i2.1095 found rates of just over 39% (Harris, 1991) and close to 46% (Bull, 1994). Although recent analyses of individual interviews have hinted at lower reply rates by current leading politicians (Bull, 2016(Bull, , 2017, the more comprehensive analysis here suggests an overall similarity with their predecessors. Individually, a broader range of reply rates was apparent across the eight political leaders. Amongst the lowest, and supportive of recent findings by Bull (2016Bull ( , 2017, were the leaders of the two main political parties in 2017: LO Corbyn at just below 30%, and PM May just short of 32%. Lowest of all across these interviews was Farron, at below 29%. Highest in their reply rates were the UKIP pair, Farage and Nuttall. Their identical reply rate figures (49.45%) showed they gave direct replies to almost half of the questions. Their apparent similarity in this aspect of interview conduct stands in stark contrast to their levels of personalisation, where, in that analysis, they were polar opposites.

Limitations
The findings herein are from two distinct periods -the general election campaigns of 2015 and 2017. Hence, it is necessary to acknowledge the importance of contemporary political context, which was not accounted for here.
Contextual influences may feature in the levels of personalisation or reply rate and, though only two years apart, the effects of these influences on the nature of the discourse are likely to vary. Indeed, the same will be true when comparing with interviews from the past. A further connected limitation relates to action orientation. Utterances, including questions and responses, should not be treated in isolation; rather, according to the conversation analysis (CA) approach, they may be understood as forms of action relevant to their specific context and oriented towards certain goals (Drew & Heritage, 1992). However, we believe the findings from this study could be used in future research to address these limitations by investigating action-orientation and the role of context in broadcast political interviews.
Future research may also consider whether there are features of questions (e.g., the presence of personalisation directed at the politician) which are more likely to be met with a response which targets the interviewer. A future study might also investigate the role of question topic in relation to personalisation. A similar analysis was conducted by  on the question-response sequences in PMQs, showing that domestic policy exchanges, compared to foreign policy, were significantly higher in personal antagonism between the leaders.
Such a study would show if question topic plays a similar role in exchanges between interviewer and politician.

Conclusions
The basis of this study was an investigation into personalised rhetoric in equivocal responses by politicians.
Featured interviews were with leaders of four UK political parties (Labour, Liberal Democrats, Conservative, and UKIP). This allowed an assessment across the mainstream political spectrum. Interviews were held during two general election campaigns (2015 and 2017) -the recency of these would facilitate an up-to-date analysis, and allow a comparison with findings from the study of interviews from around a generation ago. The method of analysis also provided an opportunity to make comparisons in reply rate -the extent to which politicians give explicit answers to interviewers' questions. Overall, the reply rate was 38%, not dissimilar to the overall rates reported for politicians from the 1980s and 1990s (Bull, 1994;Harris, 1991).
In terms of personalisation, results here were supportive of the typology devised by Waddle and Bull (2016). All equivocal responses directed personally at the interviewer were identifiable in accordance with the typology. Five of their seven categories were identified in this corpus. Most were of the critical variety; though over a quarter Journal of Social and Political Psychology 2020, Vol. 8(2), 560-581 https://doi.org/10.5964/jspp.v8i2.1095 were good-natured. The type of personalisation seemingly a tactic of former cabinet minister Peter Mandelsonrepeatedly telling the interviewer to calm down -did not feature here. Overall, 7% of equivocal responses (occasions when a full reply was not forthcoming) contained personalisation. The politician highest in personalised rhetoric here was Nigel Farage, with over 17% of his equivocal replies identified thus. However, his successor as UKIP leader, Paul Nuttall, made no such responses. The interpretation from these findings is that their position on the political spectrum is not a factor, suggesting it is more a matter of personal style. Also on an individual level, another finding of interest related to former PM David Cameron. A relative frequent user of personal attacks on opponents at PMQs , he did not exhibit a similar tendency in these interviews.
The reported effectiveness of personalisation (Bull & Mayer, 1993) -the strong likelihood of it diverting the interviewer from repeating a question that has gone unanswered -was supported by the results of this study. Bull and Mayer also reported that PM Margaret Thatcher was the politician inclined to direct personal responses at the interviewer. Results for the eight party leaders here showed only Farage to have a greater inclination for personalised equivocation than Thatcher. This finding, coupled with the now broader scope of identifying personalisation, is not evidence of increased levels by recent politicians. However, all but one of the eight leaders directed responses personally at the interviewer in lieu of a straight answer. It seems for most politicians, playing the man, in one form or another, is not beyond the pale.
Notes i) Vlaams Belang (Flemish Interest) is a Belgian nationalist party that was formed to replace Vlaams Blok (Flemish Bloc), which was dissolved in 2004 following condemnation for racism by a Belgian court (Simon-Vandenbergen, 2008).
ii) This study formed part of the Ph.D research of the first author, and an earlier version is printed in his doctoral thesis: Waddle, M. (2018). Playing the man, not the ball: Personalisation in political discourse (Doctoral thesis, University of York, York, United Kingdom). http://etheses.whiterose.ac.uk/id/eprint/22456 iii) Our study uses a content-analytic approach to address the research questions. The authors fully acknowledge the validity of alternative approaches, e.g., discursive/CA analyses (see, e.g., Hutchby & Wooffitt, 1998;Potter, 2012), and recognise the relative shortcomings of the current approach in terms of action orientation.
iv) The 2015 and 2017 elections were the most recent at the time of our analysis. v) "Paddy Backdown" was a nickname Skinner used for the then Liberal Democrats leader Paddy Ashdown. vi) Occasionally, the interviewer and politician briefly exchange small talk, typically before the start of the interview. Small talk questions like "Shall we get our mutual celebrations for the Arsenal victory out of the way first?" are not included in the analysis. This example occurred in the Peston-Corbyn interview. Both men are supporters of Arsenal Football Club.