The effect of the tenure of a coach on the relationship between the result of the match and the use of collective and individual talk
Jente Bulthuis S3856615
J.Bulthuis.2@student.rug.nl
July 2020 Master Thesis
Master in Human Resource Management Faculty of Economics and Business University of Groningen
Abstract
Lately, research has shown that collective talk has many positive effects on teams, organizations and leaders. But little is known about what causes leaders to talk collectively or individually. The current paper aims to shed light on the effect of the tenure of a coach on the relationship between the result of the match and collective and individual talk. To be able to test the hypotheses, I used content analysis. I analyzed 228 matches, using a computerized text analysis program and SPSS. The formal interviews of 18 football coaches of the Dutch Eredivisie and 20 of the Premier League were used to determine how they scored on individual and collective talk. Unlike what was expected, it appeared that more individual talk was used during interviews when the result of the coach’s team was positive. I discuss the implications for coaches, leaders and organizations concerning speaking collective. Overall, this paper serves as further evidence on the antecedents of collective and individual talk for leaders.
What Causes Leaders to Talk Collectively or Individually?
"We played the best football today. The most important thing is that the whole team felt like a family and that is why we won today". These were the words of Ajax’ trainer Ten Hag after a crucial win, making them the Eredivisie champions of 2019. In contrast, after the loss to Willem II, he said: 'Some of our players thought they could do it at 80% today, I think that is unacceptable' (FOX, 2019). These two examples demonstrate that after a win, the leader of the team (Ten Hag) speaks positively about his team. He speaks about the team being a family and playing very well, whereas in the second example, the leader speaks more about particular players instead of the team. This emphasizes how the use of collective talk in a coach’s statement differs depending on the result of the match.
To date, little is known about what causes leaders to talk collectively or individually. According to Steffens and Haslam (2013), successful leaders use more collective talk where less successful leaders use more individual talk. Holmes (1992) found that men use less collective talk than women; they tend to use words like ‘I’ and ‘me’ more often in conversations than women do. But Zimmerman, Wolf, Bock, Peham and Benecke (2013) found that people with recent problems, awful events or a loss, use more individual talk. The results of Kacewicz, Pennebaker, Davis, Jeon and Graesser (2013) show that officers with a shorter tenure tended to use more individual talk compared to officers with a longer tenure. Correspondingly, officers with a longer tenure used more collective talk compared to the short tenure officers. It seems that researchers are divided on what causes leaders to speak collectively or individually.
Building on social psychological research about what causes leaders to use collective or individual talk (Burke et al., 2007; Holmes, 1992; Kacewicz et al., 2013; Son et al., 2011; Steffens & Haslam, 2013; Zimmerman et al., 2013), I propose that the wins and losses and the tenure of a leader or coach play an essential role. Therefore, the goal of the present research is to examine the roles these variables play concerning the use of collective and individual talk by coaches. Unlike most previous studies about leaders in organizations and their way of talking (e.g., Steffens & Haslam, 2013), this study will analyze coaches of football teams. I will analyze 228 matches, using a computerized text analysis program and SPSS. The interviews, right after a match, of 18 football coaches of the Dutch Eredivisie and 20 of the Premier League will be used. To this end, I conduct a theoretical framework about the positive effects and the importance of speaking collectively for both teams and organizations as well as what the reason for talking collectively or individually could be.
Zimmerman et al., 2013) about what causes sports leaders to talk collectively or individually. Unlike previous studies (e.g. Son et al. 2011), this paper will analyze the moderating role of tenure. Hopefully, after this research I will have the answer to the question if coaches are more likely to talk collectively when their tenure as a coach is longer and vice versa.
Furthermore, this research extends previous research about momentum. Momentum is defined as a negative or positive change in physiology, cognition, effect and behavior caused by a previous event or series of events that will result in a shift in performance (Taylor & Demick, 1994). Despite a lot of research into momentum in sports, researchers seem to be divided on whether momentum is real or just an illusion (Burke, Edwards, Weigard & Weinberg, 1997). According to Vallerand, Colavecchio and Pelletier (1998), momentum in sports influences performance by both personal (motivation and skill level) and situational variables (coach and crowd behavior). This is backed up by evidence of Crust and Nesti (2006), saying that the behavior of the coach is essential for the momentum in sports. The explanation might be that when a coach uses more collective talk following a positive result, the athletes have a higher chance of winning the next game as well. Furthermore, this research has important implications for practitioners. After this research, more will be known about coaches’ use of collective and individual talk, which is essential because of the positive results of collective talk.
results, where I reflect on my theory and on assumptions I made throughout this research. The final section concludes.
Theoretical Framework
Scientific research has shown the importance of what a coach or leader is saying. Existing research recognizes the crucial role of leadership behavior of the manager in an organization because research has shown that it has a significant impact on job satisfaction and performance of employees (Koc, 2011). This applies to all leaders, managers and coaches of a club or organization. A coach is identified as a powerful agent in the sports domain, which means that at all competitive levels from professional sports to amateur sports the behavior of the coach plays an important role (Amorose & Horn, 2001). According to Smoll and Smith (2002), the coach’s behavior has a significant impact on athletes’ behavior, cognition and affective response, sense of confidence and self-determined orientation.
On the other hand, coaching behavior and attitude that is interpreted wrongly or inappropriately can lead to adverse reactions and outcomes, such as a decrease of intrinsic motivation (Amorose & Horn, 2001). Through words and actions, coaches not only have an impact on the performance of the athletes but also on the social environment. Moreover, coaches have an impact on the well-being and perspectives of the athletes and the club. Because of this, the coach is often seen as ‘the face’ of a sports club (Lyle & Cushion 2010).
This research sheds its light specifically on leaders’ collective and individual talk.
Leadership is applicable across a broad range of context (Northouse, 2019), which results in a
on all issues that affect team performance directly and is thus viewed as accountable by the board of directors (Loughead, Hardy & Eys, 2005).
Collective Talk
A coach’s words are closely monitored and therefore it is essential to know what message a coach conveys when using certain words (Borland, Burton, & Kane, 2014). First-person plurals (e.g. ‘us’, ‘we’) show a sense of belongingness, positive evaluation and cohesion of the in-group (Ruscher, 2001). Belonging to the in-group causes people and coaches to use more collective talk (Miller, Maner & Becker, 2010). The definition of collective talk is: “Communication or expression in spoken words about a group of people” (Slodov & Konya, 2015, p. 2). This communication style involves first and third person plural pronouns like ‘we’, ‘us’, ‘our’, ‘they’ & ‘them’, as expressed by the coach or leader of the team.
Several researchers have indicated that collective talk also has positive effects for the leader himself and not only for his or her followers. According to Hougaard and Carter (2018), “An ego can prevent a talented manager from becoming a great leader” (p. 1). Essentially, that is one of the reasons why the best leaders and coaches avoid using words like “I”, as well as other individual pronouns like “myself, me and mine”. In the book ‘The mind of the leader’ it becomes clear that the most effective leaders and coaches use emphatically more first-person plural pronouns, such as ‘we’, ‘us’ and ‘our’. Ineffective leaders use far less first-person plural pronouns (Hougaard & Carter, 2018).
showed that the more collective talk the leaders use, the more they are attempting to bring the audience closer to their view, and the more influential the group cohesion is (Pennebaker & Chung, 2007).
Furthermore, research from Steffens and Haslam (2013) analyzed the speeches of the official electoral campaign of prime ministerial candidates that were successful and unsuccessful. The data was drawn from all 43 Australian federal elections since 1901. They measured candidates’ use of collective talk. In 80 % of all elections, victors used more collective talk than the other not successful competitors. Victors made 61% more use of ‘us’ and ‘we’, using these pronouns once every 79 words in contrast to their other unsuccessful opponents using them only once every 126 words. This research is extending the Social Identity Theory. “Electoral endorsement is associated with leaders’ capacity to engage with, and speak on behalf of, a collective identity that is shared with followers whose energies and support they seek to mobilize” (Steffens & Haslam, 2013, p. 1). Tajfel and Turner (1979), proposed that the groups (e.g. football teams) people belong to are essential for the individual’s pride and self-esteem. They introduced the Social Identity Theory, which states that the in-group (‘us’) will distinguish themselves from the out-group ('them’) to enhance their self-image.
Moreover, in two separate studies of Ruder and Gill (1982), it has been investigated what the immediate effects of a win or loss in sports are on perceptions of cohesion. Their results show that after a win, the perception of cohesion increased on three dimensions: cohesion, sense of belonging and level of teamwork. In contrast, the perception of cohesion decreased after a loss. This indicates that perceptions of cohesion are connected to the immediate effects of a win or loss (Ruder & Gill, 1982). Based on the previously mentioned research, one can assume that a positive result of the match is associated with collective talk.
Hypothesis 1: The result of the match is associated with collective talk, such that when the result of the match is positive, it is associated with increased use of collective talk.
However, belonging to the in-group and creating the feeling of cohesion takes time. According to Tajfel and Turner (1979), one does not belong to the in-group immediately. There are cognitive processes an individual must go through to become part of the in-group, such as social identification, which is a process where individuals determine how compatible they are with a group. A process like this takes time. So, the longer the tenure of the coach or manager, the stronger his social identity, which increases the chance that the coach belongs to the in-group (Tajfel & Turner, 1979), and causes the coach to use collective talk more frequent.
lead to immediate performance change, which slows down the process of the coach joining the in-group of the team.
In another study of Kacewicz et al. (2013), researchers collected forty letters written by high-ranking officers (the officers who have been in service the longest) and lower-ranked officers (the officers who have been in service shorter) in the Iraqi military. The high-ranking officers, with a longer tenure, used more collective talk, compared to the lower-ranked officers (Kacewicz et al., 2013).
Furthermore, empirical evidence has shown that a leader with a long tenure has many positive effects on teams and organizations. Performance improves when managers and employees work together for a sustained period. The reason for this is that when there is a turnover occurs, and either the new manager or the employees need time to learn about the relationships, procedures and subcultures of a team, the performance of that team might suffer (Groysburg, Nanda, & Nohira, 2004). As a result, the tenure of coaches/ managers of a firm is associated with the quality of their performance. The longer an individual has worked with the manager or coach, the better the performance (Harris & Mcmahan, 2008).
The long tenure of a coach brings a closer relationship with the team, which leads to more use of collective talk. Tajfel and Turner (1979) claim that from the moment the relationship between the manager and players becomes closer, the manager belongs to the in-group. Additionally, long tenure of a coach and players increases the collective team identification, team learning and performance and, as a result, causes a coach to use collective talk more often. Therefore, I expect that when the result of the match is positive, a coach with a long tenure will speak more collectively than a coach with a shorter tenure will.
Figure 1
Individual Talk
Based on previous research (Hougaard & Carter, 2018; Miller, Maner & Becker, 2010; Steffens & Haslam, 2013; Son et al., 2011), it is clear that collective talk has many positive effects in contrast to individual talk. Nevertheless, leaders still use individual talk in speeches or interviews. The definition of individual talk is: “Communication or expression in spoken words about an individual” (Slodov & Konya, 2015, p. 2). This involves first, second and third singular pronouns like ‘I’, ‘me’, ‘you’, ‘he’ and ‘she’.
Like sports teams, most other organizations have also appointed leaders and managers that stand both symbolical and practical for organizations in the public eye. They are also responsible for the successes and losses of the organizations (Arnulf, Mathisen & Haerem, 2012). Two recent studies used I-talk to examine narcissism of CEOs. In the research, the authors used the indirect assessment strategy of tracking CEOs’ first-person singular pronouns (‘I’, ‘my’) use in annual reports and interviews. Results show that narcissistic CEOs tend to use more first-person singular pronouns when there were big losses for the company (Chatterjee & Hambrick, 2007).
Davis and Brock (1975) researched the use of first-person pronouns as an indication of increased objective self-awareness and performance feedback. They found an effect between individual talk and specific attention. Namely, at the moment that participants receive more attention under the ‘in the camera or mirror’ conditions, they use more individual talk. However, when participants received positive feedback in front of the camera or mirror, the amount of individual talk did decrease. In other words, at the moment that the conversation between the interviewer and the participants was about positive things about the participant, he or she started to use less individual talk (Davis & Brock, 1975). According to Borland et. (2014), media coverage of coaches and their team is unparalleled. Not surprisingly, it is crucial how coaches react in media. In their book “Sport leadership in the 21st century” it becomes clear that after a crisis, teams and coaches should be aware of reducing their individual talk. The result shows the opposite, namely that teams and coaches are using more individual talk in the media after negative events (e.g. negative media coverage) than after positive events.
the word ‘I’ more than the never-depressed participants did. Based on the previous research above, one can assume that a negative result of the match is associated with individual talk.
Hypothesis 3: The result of the match is associated with individual talk, such that when the result of the match is negative, it is associated with increased use of individual talk.
Furthermore, the research of Kacewicz et al. (2013) found that the lower-ranked officers, with a shorter tenure, tended to use more individual talk compared to the high-ranking officers, with a longer tenure. This empirical finding is backed up by the evidence of Abe (2009) that further supports the view of a decrease in use of individual talk when the tenure of a project last longer. This study applied word count strategies to examine whether the words that students described in their internship would predict the period of the internship. As one might expect, there was a decrease in use of individual talk, the longer the course of internship lasted. At the beginning of the internship the use of individual talk was significantly higher than after a longer period of the internship (Abe, 2009).
Additionally, the research of Pennebaker and Chung (2007) about the political behavior of Bin Laden and Zawahiri is showing a surprising shift of Zawahiri’s use of individual talk. This increase suggests greater insecurity and feelings of threat (Pennebaker & Chung, 2007). What is remarkable is that Zawahiri had been a leader for a shorter period compared to Bin Laden. Based on the previous research above, I expect that the relationship between a negative result of the match and the use individual talk depends on the tenure of the coach.
Figure 2
Method
To analyze the relationship between the result of the match and the collective and individual talk moderated by the tenure of the coach, desk and content analysis has been used (Holsti, 1969). My level of analysis was at the match level. I have compared matches using interviews of the coaches. Short after every match, the coach gives a brief interview. In the Eredivisie, the questions are asked by football analysts from FOX Sports, who analyzed the match themselves. In the Premier League, the questions are asked by football analysts from BBC. The items are always about the game. Example questions are: “How do you feel your team played?” and “Do you have any complaints about some players today?” The average duration of an interview is 3.07 minutes (SD = .55).
Procedure
The formal interviews of 18 football coaches of the Dutch Eredivisie and 20 of the Premier League were used to determine how they scored on individual and collective talk. Content analysis is a technique used to make inferences by objectively identifying exceptional
characteristics of messages (Holsti, 1969). In other words, content analysis is a research tool used to determine the presence of certain concepts, themes and terms within pieces of given qualitative data (i.e. text). With recent advancements in technology, the psychology of language and the computational linguistics, computerized text analyzes are increasingly reliable and efficient.
A list of the football coaches can be found in Appendix A. The interviews of the Eredivisie coaches were downloaded from the FOX Sports website. The interviews of the Premier League coaches were downloaded from the BBC website. These particular interviews were chosen because the coaches are interviewed right after the game. As literature has shown (Ruder & Gill, 1982), a difference in the use of collective and individual talk can be seen shortly after the game. The Premier League interviews used for this research are from the period between January 26, 2019, and March 9, 2020, and the Eredivisie interviews are from between January 1, 2019, and March 8, 2020. During one whole competition, there are 35 playdays. Playday 35 is the last round. The period when the post-match interviews of this research took place was around playday 25. This means that the competition was beyond the halfway point. Population and sample
For the data collection I retrieved 6 interviews of each of the 38 football coaches (38 coaches [Mage = 50.45, Min = 32, Max = 72, SD = 9.00; 100% male]). This resulted in a sample consisting of 228 matches. The football coaches’ nationalities were mostly Dutch (39.5%) followed by English (28.9%). The other nationalities that came across were German (7.9%), Portuguese (5.3%) and Spanish (5.3%). The most games were played in February (64.9%) followed by March (23.7%).
Collective and individual talk. Each of the transcripts of the interviews was analyzed using a computerized text analysis program named Linguistic Inquiry and Word Count 2015 (LIWC). The LIWC is a quantitative research tool that counts the number of words per word category in an objective and automated manner. The founders of LIWC explain “LIWC reads a given text and counts the percentage of words that reflect different social concerns, parts of a speech, thinking style and emotions” (Pennebaker, Francis, & Booth 2013, p. 88).
LIWC has a dictionary with common words and codes. The LIWC2015 dictionary is composed of 6,400 words and is used to analyze texts. Each word defines one or more codes. For example, the word ‘cried’ belongs to four-word codes: negative emotion, past tense verb, sadness and overall effect (Pennebaker, Francis, & Booth 2013). I added my dictionary in LIWC because I wanted other words and codes that were not included in the standard dictionary of LIWC. To make my dictionary, I needed to put the words and codes that I wanted to use in LIWC in excel. Every word was given code one, which stands for total talk; this is ultimately important for the sum to calculate the score for collective and individual talk, followed with a two or a three. Two stands for the collective talk and three for the individual talk (see Appendix B). The next step was to copy this excel sheet to the notepad app on my laptop, followed by changing the file extension from TXT to DIC. The last step was loading the new dictionary DIC file into LIWC, making sure that the new dictionary was switched on and the dictionary of LIWC was switched off. The dictionary is described in Table 1.
Table 1
Categories of Content Analysis
All the answers and questions received a score on individual and collective talk. LIWC gives the score in the form of a number. LIWC has a self-made sum to result in a certain score. The sum for collective talk is as follows: ‘The number of total collectively spoken words / (the
Collective talk (English)
Talking with ‘we’ or ‘us’ without
incentive
‘we’ ‘us’ ‘we’’ ‘our’
We played the best football today.
Naming just the team.
‘team’ The most important thing is that we feel like a family with the whole team, and that is what has bought us the win today.
Individual talk (English)
Talking with I of I’m 'I' 'I'' 'me.' I am not pleased with everything I saw today.
Naming just the player or the keeper. Talking 'him' or 'he.'
‘him’ ‘he’ ‘player’ ‘keeper’
He did great today. He is always in the right position to finish.
Collective talk (Dutch)
Talking with ‘wij’ or ‘ons.'
‘We’ ‘Wij’ ‘Onze’ ‘Ons’
We hebben ons aan het plan gehouden en dat levert een overwinning op. Naming just the
team
‘team’ Het team is verantwoordelijk voor deze fouten.
Individual talk (Dutch)
Talking with ‘ik’ of ‘mij.'
'Ik' 'mijn' 'mezelf.'
Ik vind het geen overtreding. Maar dat is mijn mening.
Naming just the ‘speler’ or the keeper.
'Hij' 'speler' 'hem' 'keeper.'
total number of spoken words /100)’. So, it was calculated how many per cent of all words were collective. The length of the interview was also taken into account in the calculation, which was important because a longer interview can increase the amount of collective or individual words spoken. The sum for individual talk was the almost the same: ‘The number of total individually spoken words / (the total number of spoken words /100)’. E.g. Brendan Rodgers’ number of total collectively spoken words is 9. His total amount of words spoken is 263. The number of collective talk = 9 / (263/100) = 3.42. The same applied to the individual talk sum. In this way, I received information about the collective and individual talk used in all the questions and responding answers.
To keep the interviews as valid as possible, I needed to make sure that no mistakes were made with the translation of measuring instruments. According to Beaton, Bombardier, Guilleman and Ferraz (2000), a simple single translation of an interview is not good enough. A good interview or questionnaire translation consists of a process of at least two or three people who perform independent forward-translations and independent backward-translations. For this reason, I decided to make both an English and a Dutch dictionary. As a result, the interviews could be analyzed in the same language, and no translation mistakes could be made.
To rate the individual talk for the English interviews, the variables that were used were ‘I’, ‘he’, ‘player’, ‘keeper’, ‘I’, ‘me’ and ‘him’. The variables that were used for the Dutch interviews were ‘ik’, ‘hij’, ‘speler’, ‘mij’, ‘mijzelf’, ‘hem’ and ‘keeper’. Those variables were based on previous research (Pennebaker & Chung, 2007; Ruscher, 2001; Steffens & Haslam, 2013) that used the same personal pronouns in their study to investigate collective and individual talk. The whole dictionary that I implemented in LIWC can be found in Appendix B. In both dictionaries ‘the player’ and ‘de speler’ is implemented. Because it is impossible to place all players in the dictionary by name, I decided that when a trainer mentioned a player, this would be quoted in the interview as ‘the player’ and not the particular name. Naming a specific player was related to individual talk because discussing and picking out one player instead of the team is a deliberate act of individual talk. For all the 228 matches the collective and individual talk was rated (Collective talk [Mscorecollectivetalk = 4.35, SD = 1.74]; Individual talk [Mscoreindividualtalk = 4.22, SD = 1.85].
Result of the match. The result of the match was operationalized in five different ways. For each interview, the number of goals for and against was kept. (Goals for [Mgoalsfor = 1.27
SD = 1.23, Min = 0, Max = 5.00]; Goals against [Mgoalsagainst = 1.27 SD = 1.23, Min = 0, Max
= 5.00]). Furthermore, it was noted whether it concerned a won (84 times), lost (84 times) or a tied match (60 times). In conclusion, the goal difference after every match was noted (Goal difference [Mgoalsdifference = 0, SD = 1.76,, Min = -4.00, Max = 4.00]). This shows symmetry in my data since every goal for/win is a goal against/loss for the other team. More collective talk was expected to be used in interviews after a game with a positive outcome (A win, goals for and goal difference). More individual talk was expected to be used in interviews after a game with a bad outcome (A loss, goals against and goal difference).
the coach was kept in the number of days and was updated every week. (Tenure of the coach [Mtenure = 575.96, Min = 20, Max = 2700, SD = 645.87]).
Control variables. Six control variables were used so that alternative explanations could be excluded and the effect would be isolated. Swaen (2017) claims that when adding control variables in studies, the results will be more accurate, and according to Becker, Atinc, Breaugh, Carlson and Edwards (2016) it is crucial to control for variables that are significantly correlating with the model.
Interviewer questions. I included the interviewers’ questions in my analysis. For
example, the answers and the questions of the interviewer, were analyzed separately on individual talk and collective talk. However, the documents were linked together, so I knew which questions belonged to which answers. It is essential to have this variable in my research because, for example, a high amount of questions about particular players can influence a coach’s score on individual talk. If a strong relationship between the result of the match and collective or individual talk is found, but no control variables are used, the result is not accurate, since an alternative explanation, such as the interviewer's questions, may apply. The questions asked by the interviewer were rated (Collective talk [Mscorecollectivetalk = 0.66, SD = 0.84]; Individual talk [Mscoreindividualtalk = 3.31, SD = 2.25]). Additionally, in my data, I found that a higher amount of individual talk of the interviewer is associated with coaches that talk more individual (r = .30, p = .00), which may mean that the coaches are influenced in their talk by the questions that are asked. This can influence the research question.
Age. The second control variable was age. Dimec and Kajtna (2009) did research about
I expected that older coaches use more individual talk in contrast to younger coaches. That is the reason I controlled for this variable. Results showed no effect between the age of the coach and the amount of collective or individual talk (r = -.02, p = .92; r = .77, p = .17).
Nationality. I included the nationality of the coach. It might seem that tenure moderates
the relationship between the result of the match and collective talk. Still, it could be that shorter tenure coaches often have different nationality and that culture influences team cohesion. According to Butter (2018), this can be explained by the fact that usually after a short period, coaches with a different nationalities do not fit well with the club. This is mainly because the coach does not thoroughly study the culture of the country and because of the language barrier. Therefore, in a correlation between the result of the match and collective and individual talk, the coach’s nationality can have a disruptive effect. According to Na and Choi (2009), the use of individual talk is dependent on one's cultural background. Therefore, in a correlation between the result of the match and collective or individual talk, there may be a difference in nationality.
It seems that the coaches with a Dutch ‘nationality’ and coaches from the Eredivisie are associated with less ‘collective talk’ (r = -.55, p = .00; r = -.39, p = .00) and coaches with an English ‘nationality’ and coaches from the Premier League are associated with more ‘collective talk’ (r = .31, p = .00; r = .39, p = .00). These correlations showed that there may be a difference in culture concerning the use of collective talk. Because I took Dutch and English coaches, I could look for these differences, which made it easier to generalize my research. This also enabled me to come up with a better answer to my research question, which is about coaches in general.
Ranking. The other control variable was the ranking of the team. This data was also
ranking could have affected the outcome. It showed that a lower ranking is associated with more collective talk (r = .22, p = .00). These correlations showed that there may be a difference in ranking for collective talk. Because I took all rankings from high to low, I could look for these differences, which enabled me to give a better answer on my research questions about coaches from all rankings.
Month and playday. The last control variables were the month in which the match was
played and the playday of the competition. This data was retrieved from the Eredivisie and the Premier League websites. The most critical months are mainly just before and after the winter break, as you approach the end. The same goes for the playday. In those critical months, a coach can have greater insecurity and feelings of threat. This can increase the individual talk of a coach (Pennebaker & Chung, 2007). Results did show an effect between the playday and the talk of a coach, but not as expected. Namely, there was an effect between the playday and the collective talk of the coach (r = .30, p = .02), which means that the later in the season, the higher the amount of collective talk. So it was indeed essential to control for these variables.
Results
Descriptive Statistics
This implies that a low rank is associated with coaches with an English nationality. This indicates that there is a difference between Dutch and English coaches which is critical to take into account for further findings because my research question is about coaches in general.
Therefore, I made an extra dummy to test this further. I took Dutch coaches and the Eredivisie together as a variable, named 'NED_ERE', and English coaches and the Premier League as a variable, named 'ENG_PRE', and this made the effect even stronger. The direction of the relationship between ‘ranking and ‘NED_ERE’ is negative ( r = -.37 p = .03) and with ‘ENG_PRE’ it is still positive ( r = .32 p = .00). This indicates that there might be a difference between culture and competition in different countries. This is important for the research question because a good result of the match is therefore more common among Dutch and non-English coaches, which in turn can affect collective and individual talk. Another interesting statically significant linear relationship is the ‘individual talk from the interviewer’ and the ‘ranking’ ( r = -.14 p = .03). This suggests that the higher the ranking, the greater the individual talk from the interviewer. This is an interesting relationship because it gives insight into the sample; it indicates that other variables influence the interviewer’s questions. This means that because a team is higher in the ranking, the more individual questions are asked, which increases the individual talk from the coach.
Furthermore, there is a positive statically significant linear relationship between ‘goals for’ and ‘individual talk of the interviewer’ (r = .18, p = .04). This indicates that this control variable can have a disrupted effect on the effect between goals for and individual talk of the coach and is therefore essential for the research question. This finding will be further discussed in the discussion. Additionally, a relationship between a 'win' and 'individual talk' is found. Namely, when the result of the match is a ‘win’, increased use of ‘individual talk’ of the coach appears (r = .16, p = .02). These variables tend to increase together (i.e., the more goals, the more individual talk). This indicates that a positive result of the match is not associated with more collective talk as expected, because, it turns out to be associated with individual talk. This finding will be further discussed in the discussion.
Some relationships already give some information about the hypothesis. These relationships will be discussed in detail in the discussion. First of all, ‘individual talk’ and ‘goals for’ have a statically significant linear relationship (r = .12, p = .04). The direction of the relationship is positive, meaning that these variables tend to increase together (i.e., more ‘individual talk’ is associated with more ‘goals for’), and also tend to decrease together (i.e., less ‘individual talk’ is associated with less ‘goals for’). Another interesting relationship is that the ‘collective talk’ and ‘goals for’ have a statically significant linear relationship (r = -.41, p = 0.03). The direction of the relationship is negative, meaning that if one variable increases, the other variable decreases (i.e., more ‘goals for’ is associated with less ‘collective talk’). The relationship between ‘win’ and ‘collective talk’ is negative (r = -.14, p = .04). There is no interesting significant effect between tenure and the dependent and independent variable.
Assumptions. To test my hypothesis, I will conduct a regression analysis. Several assumptions should be met, such as the variables should be normally distributed, and the variables should not have any outliers.
Outliers. There is a small number of coaches with an extraordinarily high or low score of individual or collective talk. There are four Z-scores of individual talk and three of collective talk, that means that those four scores have three or more SDs above or below the average, which makes it an extremely high or low score. Because there are seven outliers, the analysis needs to be done twice; once with outliers and once without. I also looked at Cook's Distance. Because both collective talk (Min = 0, Max = .101) and individual talk (Min = 0, Max =.060) were not above a maximum of one, the Cook’s distance is normal. Since deleting the outliers had no significant effect, and the Cook's distance was normal, I decided to keep the outliers from the analysis and continue with the regression analysis.
Figure 3
Figure 4
Histogram Score Collective Talk
Table 2 Descriptive Statistics M SD 1 2 (1) 2(2) 2(3) 3 4(1) 4(2) 4(3) 5 6 7 8(1) 8(2) 9 10 11 12 13 14 15(1) 15(2) 15(3) Control variables 1. Age 50.45 9.00 1 2. Nationality 2.1 Netherlands .39 .49 -.010 1 2.2 England .30 .45 .077 -.515** 1 2.3 Other .32 .47 -.065 -.549** -434** 1 3. Ranking 10.01 5.55 -.067 -.180** .252** -0.57 1 4. Month 4.1 January .11 .32 -.035 -.233** .136* .112 .017 1 4.2 February .65 .48 -.018 .048 .003 -.054 .093 -.488** 1 4.3 March .23 .43 .047 .120 -.105 -.023 -.116 -.200** -.758** 1 5. Playday 24.92 2.30 .015 -.499* .287** .245** .058 -.300** -.209** .459** 1 6. Coltalk interviewer -.66 -.84 -.017 .104 -.137* .025 .008 .016 .068 -0.88 -.042 1 7. Indtalk interviewer 3.31 2.25 .151* .125 -.100 -.034 -136* -.090 .030 .034 -.053 -.032 1 8. League 8.1 Eredivisie .47 .50 .000 .851** -.489** -.418** -.090 -.285** .108 .091 -.587** .090 .123 1 8.2 Premier League .53 .50 .000 -.851** .489** .418** .090 .285** -.108 -.091 .587** -.090 -.123 -1.000 1 Model variables 9. Goals for 1.27 1.23 .011 .106 -.141* .026 -.235** -.046 -.002 .036 .022 .021 .179** .069 -.069 1 10. Goals against 1.27 1.23 .075 .055 .00 -.058 .198** -.046 -.002 .036 .022 .080 -.102 .069 -.069 -0.14 1 11. Tenure 575.96 645.90 -.124 -.223** -.263** -.023 -.013 .074 -.027 -0.25 -.184** -.036 -.206** -.287** .287** -.032 -.058 1 12. Collective talk 4.35 1.74 -.019 -.548** .314** .270** .219** .045 .019 -.055 .296** -.164* -.202** -.393** .393** -.141* .002 .077 1 13. Individual talk 4.22 1.85 .092 .022 -.007 -.016 -.080 .055 -.064 .031 -.001 .077 .298** -.054 .054 .136* -.025 -.073 .444** 1 14. Goaldifference .00 1.76 -.045 .036 -.099 .059 -.304** .000 .000 .000 .000 .071 .197** .000 .000 .712** -.712** .018 -.100 .113 1
15. Result of the Match
15.1 Win .37 .48 .049 .090 -.167* .068 -.309** -.074 .009 .045 .022 .074 .157* .058 -0.58 .659** -.450** .-87 -.136* .018 .778** 1
15.2 Los .37 .48 -.020 .016 .014 -.030 .176** -.074 .009 .045 .022 -.035 -.120 .058 -.058 -.450** .659** -.026 .009 -.068 -.778** -.583** 1
15.3 Tie .26 .44 -.031 -.116 .168* -.042 .145* .162* -.020 -.099 -.049 -.043 -.040 -.128 .128 -.229** -.229** -.068 .138** .045 .000 -.456** -.456** 1
Analysis
To test my hypotheses, I conducted three regressions. (1) A regression with all control variable, (2) a regression with tenure as moderator included, without control variables, and (3) a regression with tenure as moderator and with all control variables. All the regressions are repeated for the two dependent variables and the five independent variables (Table 3; Goals for, Table 4; goals against, Table 5; goal difference, Table 6; Win and Table 7: Loss).
Hypothesis 1. The result of the match is associated with 'collective talk', such that: When
the result of the match is positive, it is associated with increased use of collective talk. Table 2
displays the descriptive statistics of the Pearson correlation test that I conducted to test this hypothesis. A positive result of the match is tested through three different variables; ‘goals for’, ‘goal difference’, and ‘win’. There were two significant relationships, namely that ‘goals for’ is negatively correlated with ‘collective talk’ (r = -.141 p = .034) and the same applies to ‘win’ and ‘collective talk’ ( r = -.136 p = .041). This contradicts my prediction of a positive relationship between ‘goals for’, ‘win’ and ‘collective talk’. The result of ‘goal difference’ and ‘collective talk’ show no significant effect of a correlation between a positive result of the match and the use of collective talk ( r = -.100 p = .131). Therefore, it is clear that there is no support for this hypothesis. Results showed a surprising significant negative correlation between a positive result of the match and the use of collective talk. Model 1 in Tables 3, 5 and 6 shows that when controlling for the control variables, the coefficients decreased, and the relationship was not significant anymore.
Hypothesis 2. The relationship between a positive result of the match and the use of
collective talk is stronger when the tenure of the coach is long rather than short. To test
tenure. First, I conducted my analysis without my control variables. Against my predictions, Tables 3, 5 and 6 (model 2) revealed no effect of the moderator (Table 3(2) B = .070, SE = .129; Table 5(2) B = .087, SE = .127; Table 6(2) B = .073 SE = .115). Therefore, it is clear that there is no support for this hypothesis. Model 3 in Tables 3, 5 and 6 shows that when controlling for the control variables, the coefficients decreased a bit.
Hypothesis 3. The result of the match is associated with ‘individual talk’, such that: When
the result of the match is negative, it is associated with the use of increased individual talk.
Table 2 displays the descriptive statistics of the Pearson correlation test that I conducted to test this hypothesis. A negative result of the match was tested through three different variables; ‘goals against’, ‘goal difference’ and ‘loss’. The analysis revealed no main effect of a negative result of the match. Surprisingly, it was the exact opposite. There was a positive significant relationship between ‘goals for’ and ‘individual talk’ (r = .12, p = .04). The direction of the relationship is positive, meaning that these variables tend to increase together (i.e., more ‘individual talk’ is associated with more ‘goals for’). For this hypothesis, I also estimated a regression model in which I controlled for my control variables; the coefficients decreased, and the relationship was not significant anymore, see Tables 4, 5 and 7 (model 4). None of the other two variables ‘goal difference’ and ‘loss’ showed a significant effect of a correlation between a negative result of the match and the use of individual talk (r = -.068 p = .309; r = -.025 p = .707). Therefore, it is clear that there is no support for this hypothesis.
Hypothesis 4. The relationship between a negative result of the match and the use of
individual talk is stronger when the tenure of the coach is short rather than long. To test this
B =.032, SE = .144; Table 5(5) B = -.203, SE = .134; Table 7(5) B = .184, SE = .128). Only a marginally significant effect for goals for and tenure was found, see Model 5, Table 3. When controlling for the control variables, the coefficients decreased, and the relationship was not significant anymore, see Table 3, Model 5. Therefore, results show that there is no support for this hypothesis.
Table 3
Goals For
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Collective Individual Variables B (SE) B (SE) B (SE) B (SE) B (SE) B (SE) Intercept 4.88(.192)*** 4.36(.115)*** 4.88(.193)*** 3.30(.240)*** 4.21(.121)*** 3.319(.240)*** Goals for -.080(.099) -.236(.115)** -.079(.099) .165(.123) .231(.122)* .153(.123) Age -.017(.097) -.017(.097) .076(.121) .081(.120) Netherlands -1.21(.188)*** -1.21(.188)*** .369(.233) .353(.232) English .045(.117) .045(.117) .063(.145) .063(.145) Ranking .121(.117) .121.(.103) -.008(.128) -.006(.127) January .111(.191) .111(.191) -.141(.238) -.135(.237) February .153(.150) .153(.151) -.218(.188) -.223(.187) Playday .308(.201) .308(.201) -.273(.250) -.276(.250)
Collective talk Interviewer -.261(.116)** -.259(.117)** .228(.145) .200(.145) Individual talk Interviewer -.108(.044)** -.108(.044)** .233(.055)*** .230(.055)*** Eredivisie .602(.234)** .601(.235)** -.676(.292)** -.656(.291)**
Tenure .127(.115) -.097(.103) -.128(.122) -.080(.127)
Goals for x Tenure .070(.129) .011(.107) -.247(.136)* -.210(.132)
R² .376 .026 .376 .139 .038 .149
R adjusted .341 .091
F 10.80*** 2.02 9.92*** 2.89*** 2.91** 2.88***
Table 4
Goals Against
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Collective Individual Variables B (SE) B (SE) B (SE) B (SE) B (SE) B (SE) Intercept 4.91(.193)*** 4.35(.116)*** 4.92(.194)*** 3.25(.241)*** 4.22(.123)*** 3.25(.243)*** Goals against -.051(.098) .008(.117) -.045(.099) .051(.123) -.052(.124) .051(.124) Age -.011(.097) -.004(.098) .069(.122) .069(.123) Netherlands -1.21(.187)*** -1.22(.188)*** .377(.234) .376(.235) English .050(.116) .050(.116) .051(.146) .051(.146) Ranking .147(.103) .144(.103) -.052(.128) -.052(.129) January .103(.190) .104(.191) -.120(.238) -.120(.239) February .146(.150) .153(.151) -.200(.188) -.200(.189) Playday .298(.200) .299(.201) -.244(.251) -.244(.251)
Collective talk Interviewer -.267(.117)** -.268(.117)** .234(146) .234(146) Individual talk Interviewer -.116(.044)*** -.117(.044)*** .246(.055)*** .246(.055)*** Eredivisie .600(.235)** .604(.235)** -.666(.293)** -.666(.294)**
Tenure .131(.117) -.094(.103) -.135(.124) -.075(.129)
Goals against x Tenure -.046(.135) .082(.112) .032(.144) .006(.140)
R² .375 .006 .377 .132 .006 .132 R adjusted .340 .084 F 10.75*** .484 9.94*** 2.73** .480 2.51** Notes: N = 228; *p <.10, **p <.05, ***p <.01. Table 5 Goal Difference
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Age -.018(.097) -.014(.098) .081(.122) .094(.122) Netherlands -1.22(.187)*** -1.22(.188)*** .382(.243) .360(.234) English .051(.117) .051(.117) .053(.146) .053(.146) Ranking .131(.105) .131(.105) -.017(.131) -.018(.131) January .100(.190) .101(191) -.120(.238) -.116(.238) February .145(.150) .146(.151) -.202(.188) -.197(.188) Playday .291(.200) .291(.200) -.240(.250) -.241(.250)
Collective talk Interviewer -.259(.116)** -.263(.117)** .220(.146) .204(.146) Individual talk Interviewer -.111(.045)** -.113(.045)** .237(.056)*** .232(.056)*** Eredivisie .593(.234)** .598(.235)** -.658(.293)** -.642(.293)**
Tenure .132(.116) -.095(.103) -.128(.122) -.074(.128)
Goals Difference x Tenure .087(.127) -.043(.105) -.203(.134) -.153(.131)
R² .374 .018 .374 .133 .028 .138 R adjusted .339 .085 F 10.72*** 1.40 9.86*** 2.76** 2.18* 2.65** Notes: N = 228; *p <.10, **p <.05, ***p <.01. Table 6 Win
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Collective Individual Variables B (SE) B (SE) B (SE) B (SE) B (SE) B (SE) Intercept 4.89(.193)*** 4.35(.115)*** 4.90(.196)*** 3.24(.241)*** 4.23(.123)*** 3.26(.244)*** Win -.054(.102) -.254(.115)** -.050(.102) -.076(.127) .050(.124) -.073(.128) Age -.014(.097) -.011(.097) .077(.121) .079(.122) Netherlands -1.22(.187)*** -1.22(.188)*** .380(.234) .375(.235) English .044(.117) .043(.117) .040(.147) .040(.147) Ranking .123(.104) .122(.104) -.061(.130) -.061(.131) January .101(.190) .102(.191) -.113(.238) -.113(.238) February .147(.150) .147(.151) -.193(.188) -.193(.188) Playday .297(.200) .292(.201) -.227(.250) -.231(.251)
Tenure .153(.115) -.090(.104) -.136(.124) -.066(.130) Win x Tenure .073(.115) -.048(.096) -.096(.123) -.038(.120) R² .375 .028 .376 .133 .009 .134 R adjusted .340 .085 F 10.75*** 2.16* 9.91*** 2.75*** .649 2.54** Notes: N = 228; *p <.10, **p <.05, ***p <.01. Table 7 Loss
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Collective Individual Variables B (SE) B (SE) B (SE) B (SE) B (SE) B (SE) Intercept 4.92(.192)*** 4.35(.116) 4.93(.194)*** 3.27(.241)*** 4.22(.122) 3.29(.242)*** Loss -.070(.097) .019(.116) -.069(.097) -.017(.121) -.126(.122) -.015(.121) Age -.016(.097) -.012(.097) .074(.121) .081(.121) Netherlands -1.22(.187)*** -1.23(.188)*** .380(.234) .363(.235) English .051(.116) .052(.117) .050(.146) .051(.146) Ranking .148(.102) .142(.103) -.038(.128) -.051(.128) January .094(.190) .098(.191) -.117(.238) -.109(.238) February .141(.150) .148(.151) -.199(.188) -.184(.189) Playday .292(.200) .299(.200) -.235(.250) -.222(.250)
Discussion
Unlike what was expected, it appeared that more individual talk was used during interviews when the result of the coach’s team was positive. This challenges the study of Chatterjee and Hambrick (2007) and Pennebaker and Chung (2007), who said that leaders tend to use more first-person singular pronouns when there were big losses in an organization and more third-person plural pronouns after an important event. A possible explanation for the fact that I did not find the same results is that their research was not about sports. It is plausible that in business the findings are different and therefore their research cannot be extended into sports.
This also contradicts the evidence of Davis and Brock (1975) and Borland et al. (2014). They claim that the coaches use less individual talk in the media after positive events than after negative events. An explanation for this result could be that with a positive result of the match, specific players have scored certain goals that are always worth discussing afterwards. This theory is backed up with the view of Leinemann and Baikaltseva (2006), who stated that journalists are always looking for the most exciting story which, in sports, often concerns personal matters. This can explain that after a win a lot of individual questions are asked. However, when controlling for the control variables, the relationship between a game in which a team scores many goals and the coach talking individually was not significant anymore. A reason for this effect could be that some control variables have a disrupted effect. For example, Leinemann and Baikaltseva (2006) claim that more individual questions are asked after a game with many goals.
authors think that the amount of exposure differs between cultures, such that one’s cultural orientation determines one’s choice between individual and collective talk. They show that Korean participants chose the use of collective talk more than the American participants did. They believe that the Korean participant’s choice was influenced by the Asian culture. This might also be the case in The Netherlands and England. According to Hofstede (1980), the culture of the Netherlands and England shows more signs of individualism than collectivism. In reality, there can be quite some differences within countries. This might be the case in England.
Even though the findings are not significant, the present research gives insight for people who want to replicate the study, namely that the coaches answer the questions that have been asked (see the significant effect of the individual- talk of the interviewer). However, results show that when the interviewer asked more questions about the collective, the coach’s answers were less about the collective. One reason for this can be that the questions about the collective were focused more on a situation rather than the actual collective. An example of this is the following question: “Perhaps we can have a look at the scene that we saw on the pitch at the
final whistle, because your team was exhausted. Shall we see for a moment?” The collective
score for this question is very high, while the question did not concern the actual collective. Theoretical implications
not universally affect a leader's use of collective and individual talk. It is plausible that there are significant differences between sports and business. The result of the match cannot be compared to losses and wins in organizations. This can be included on the theoretical level to expand the theory. Unfortunately I cannot extend previous research of momentum (Burke et al., 1997) on whether it is real or just an illusion. Results show no relationship between collective or individual talk and winning or losing the next game.
Furthermore, the findings showed that how coaches talk after a game in which their team has scored many goals does not seem to differ between coaches who have just started and those who have been working at a club for a long time. In the situation of a football coach who has been the coach of one team for a longer period, the effects of Social Identity Theory (Tajfel & Turner, 1979) is not perceptible. It is clear that when a coach or manager has been working with the same team for a longer period, the social effects of becoming part of the in-group and the increased use of collective talk did not appear. It may be that the theory is only applicable on members of the same level and that it does not apply to a leader or coach. This could be included on the theoretical level to expand the Social Identity Theory.
Moreover, coaches from the Netherlands talk less collectively and coaches from England more collectively. This effect could be a cultural difference. This outcome is consistent with the previously mentioned theory of Hofstede (1980) that stated that European countries are more individualistic. It may be that there is a difference between European countries. This theoretical finding could be included to expand the theory. This finding is strengthened by the Turnhout (2017), who says that Dutch people are direct and straightforward. According to Hofstede (1980), this fits with an individualistic culture. In individualistic cultures, it is appreciated that you express your own opinion and the communication is direct, explicit and honest. This can be the reason that Dutch coaches talk collectively less frequent. It is a necessary theoretical implication that there is a specific culture difference in the interviews. It could, however, also be a league difference. The English Premier League is the most powerful league in the world. They make the most significant impact on the international stage (the global area, television-rights and the dominant position in Europe, etc.). The fact is, coaches in the English Premier League are under a lot more pressure (Pilger, 2014). As a result, they might be more closely monitored by external parties. Therefore, their directness and honesty decreases, which can lead to less use of individual talk. This can be another reason why there is a reasonable difference between the Eredivisie and the Premier League.
Practical implications
individual and collective talk by a coach differs per country and league. Clubs can take this observation into account when hiring a new coach from a different country.
The results reported here also offer another serious practical implication for organizations or sports clubs that strive to performance improvement, cohesion and positive emotions. The findings suggest that organizations, leaders or coaches should pay more attention to the perceived speaking words in front of a camera. Although speaking collectively regarding the individual questions is a tricky thing (Leinemann & Baikaltseva, 2006), organizations and leaders that strive for the positive effects of collective talk should ensure that the leaders talk collective during interviews. Organizations or leaders could increase such a way of speaking by, for example, media training. Results showed that the tenure played no role in this matter. Therefore every football coach, with either a short or long tenure, should follow the media training.
Limitations and Directions for Future Research
This research faces a couple of limitations. Those limitations have to be kept in mind when interpreting the results and are essential for further studies. It might be that I interpreted the theory (Chatterjee & Hambrick, 2007; Rabbie & Horwitz, 1969; Steffens & Haslam, 2013; Tajfel & Turner, 1979) otherwise or even in a wrong way. It could be that the research I found cannot be extended towards football coaches because none of the literature focused on sports. For future research, the view around those topics may be reviewed more independently and extensively.
Division from the Netherlands and the Championship from England. If possible, concerning the language barrier, I would even recommend expanding to leagues from other countries.
Furthermore, I would suggest using more interviews per coach. Due to time constraints, six interviews per coach were used, but for further research, I would aim for at least ten interviews per coach. Moreover, there are limitations to the recorded interviews. The interviews were conducted by various reporters, which means that questions are asked in different ways. One asks more critical questions, the other more positive or direct questions. These can subconsciously lead to more individual or collective talk. For further research, it would be interesting to use the ‘speeches’ that a coach gives to his players during the break and after the game. It could be a limitation that the interviews that I used are given in front of a camera and used on television and are not explicitly aimed at the players, making the coaches less aware of their way of talking. This is backed up by evidence of Azzoni (2018), who claims that the best moment for coaches to influence their teams are in the closed dressing room. The conversations in the dressing room are filled with deep meaning. Every word that is said there contributes to the win (Azzoni, 2018).
A third limitation could be the measurement of the collective and individual talk. The dictionary chosen to filter on collective and individual talk was not equally defined. The word ‘I’ was not well chosen, because ‘I’ is overrated for the use of individual talk. Coaches often use the word ‘I’ before a statement about the collective. For example: “I think we did a
wonderful job today”. "I am very proud of the team". For future research, I would suggest
about why this could be the case, but it would be interesting to know more about why this occurs.
Conclusion
References
Abe, J. A. A. (2009). Words that predict outstanding performance. Journal of Research in
Personality, 43(3), 528–531. https://doi.org/10.1016/j.jrp.2009.01.010
Amorose, A. J., & Horn, T. S. (2001). Pre- to Post-Season Changes in the Intrinsic Motivation of First Year College Athletes: Relationships with Coaching Behavior and Scholarship Status. Journal of Applied Sport Psychology, 13(4), 355–373.
https://doi.org/10.1080/104132001753226247
Arnulf, J. K., Mathisen, J. E., & Hærem, T. (2012). Heroic leadership illusions in football teams: Rationality, decision making and noise-signal ratio in the firing of football managers. Leadership, 8(2), 169–185. https://doi.org/10.1177/1742715011420315. Azzoni, T. (2018, May 23). Zidane's Dressing Room Management Key To Madrid's Success.
Consulted from: https://apnews.com/00733b7407da48caa4271bc4ef1dcaf2/Zidane's-dressing-room-management-key-to-Madrid's-success.
BBC, (2020). Premier League reactions. Consulted from: https://www.bbc.com/sport/av/football/51525580
Beaton, D. E., Bombardier, C., Guillemin, F., & Ferraz, M. B. (2000). Guidelines for the Process of Cross-Cultural Adaptation of Self-Report Measures. Spine, 25(24), 3186– 3191. https://doi.org/10.1097/00007632-200012150-00014
Borland, J. F. B., Burton, L. J., & Kane, G. M. (2014). Sport Leadership in the 21st Century. Burlington, United States: JONES & BARTLETT LEARNING.
Boons, K. (2017). Factcheck: ‘Trainers worden steeds sneller ontslagen’. Factory.
https://factory.fhj.nl/factcheck-trainers-worden-steeds-sneller-ontslagen/
Burke, K. L., Edwards, T. C., Weigand, D. A., & Weinberg, R. S. (1997). Momentum in sport: A real or illusionary phenomenon for spectators. International Journal of Sports
Psychology, 28, 79-97.
Butter, J. (2018). In het buitenland zit niemand meer te wachten op Nederlandse voetbaltrainer. Consulted from https://www.transfermarkt.nl/.
Carron, A. V., Eys, M. A., & Burke, S. M. (2007). Team cohesion: Nature, correlates, and development. In: S. Jowett, & D. Lavallee (eds). Social psychology in sports 91-101. Champaign, IL: Human Kinetics.
Chatterjee, A., & Hambrick, D. C. (2007). It's All About Me: Narcissistic Chief Executive Officers and Their Effects on Company Strategy and Performance. Administrative
Science Quarterly, 52(3), 351–386. https://doi.org/10.2189/asqu.52.3.351
Carolin, J. (2016). The Barefoot Leader. Bloomington, Indiana: Xlibris AU.
Crust, L., & Nesti, M. (2006). A review of psychological momentum in sports: why qualitative research is needed. Journal of Sport Psychology, 8(1), 1-6.
Cushion, C., & Lyle, J. (2010). Sport Coaching: professionalization and practice. Churchill
Livingstone, Edinburgh.
Davis, D., & Brock, T. C. (1975). Use of first person pronouns as a function of increased objective self-awareness and performance feedback. Journal of Experimental Social
Dimec, T., & kajtna, T. (2009). Psychological characteristics of younger and older coaches.
Kinesiology, 41, 172–180.
Dios Tena, J., & Forrest, D. (2007). Within-season dismissal of football coaches: Statistical analysis of causes and consequences. European Journal of Operational
Research, 181(1), 362–373. https://doi.org/10.1016/j.ejor.2006.05.024
Dictionary, CE (1979). Complete & Unabridged 2012 Digital Edition© William Collins Sons & Co. Ltd. 1979, 1986© HarperCollins.
FoxSports Interviews (2020). Consulted from
https://www.foxsports.nl/video/filter/interviews/
Groysberg, B., Nanda, A., & Nohria, N. (2004). The risky business of hiring stars. Harvard
business review, 82(5), 92-101.
Hakan Koç. (2011). The impact of managers’ leadership behaviors on job satisfaction and performance of employees. African Journal of business management, 5(30). https://doi.org/10.5897/ajbm10.1229
Harris, C. M., & Mcmahan, G. C. (2008). An empirical investigation of human capital and performance: NCAA Men’s Basketball teams. Academy of management proceedings,
2008(1), 1-6. https://doi.org/10.5465/ambpp.2008.33622338
Hofstede, G. (1980). Culture and Organizations. International Studies of Management &
Organization, 10(4), 15–41. https://doi.org/10.1080/00208825.1980.11656300
Holmes, J. (1992). Women’s Talk in Public Contexts. Discourse & Society, 3(2), 131–150. https://doi.org/10.1177/0957926592003002001
Holmes, P. (2010). Win or Go Home: Why College Football Coaches Get Fired. Journal of
Holsti, O. R. (1969). Content analysis for the social sciences and humanities. Reading. MA:
Addison-Wesley (Content analysis).
Hougaard, R., & Carter, J. (2018). The Mind of the Leader. Amsterdam, Nederland: Reed Business Education.
Kacewicz, E., Pennebaker, J. W., Davis, M., Jeon, M., & Graesser, A. C. (2013). Pronoun Use Reflects Standings in Social Hierarchies. Journal of Language and Social
Psychology, 33(2), 125–143. https://doi.org/10.1177/0261927x13502654
Leinemann, R., & Baikaltseva, E. (2006). How to Manage a Successful Press Conference. Hampshire, England: Gower.
Loughead, T. M., & Hardy, J. (2005). An examination of coach and peer leader behaviors in sport. Psychology of Sport and Exercise, 6(3), 303–312.
https://doi.org/10.1016/j.psychsport.2004.02.001
McTeer, W. G., White, P. G., & Persad, S. (1995). Manager/coach mid-season replacement and team performance in professional team sports. Journal of Sport Behavior, 18(1), 58-68.
Mielke, D. (2007). Coaching Experience, Playing Experience and Coaching
Tenure. International Journal of Sports Science & Coaching, 2(2), 105–108. https://doi.org/10.1260/174795407781394293
Miller, S. L., Maner, J. K., & Becker, D. V. (2010). Self-protective biases in group categorization: Threat cues shape the psychological boundary between “us” and “them”. Journal of Personality and Social Psychology, 99(1), 62–77.
Na, J., & Choi, I. (2009). Culture and First-Person Pronouns. Personality and Social
Psychology Bulletin, 35(11), 1492–1499. https://doi.org/10.1177/0146167209343810
Northouse, PG (2019). Leadership: Theory and practice. Glasgow: Sage.
Pennebaker, J. W., & Chung, C.K. (2007). Computerized Text Analysis of Al-Qaeda Transcripts. In K. Krippendorff & M.A. Bock (Eds.), A content analysis reader (pp.453-465). Thousand Oaks, CA: Sage.
Pennebaker, J. W., & Francis, M. E., & Booth, R. J. (2013). Linguistic inquiry and word
count: LIWC 2001 manual. Mahwah, NJ: Erlbaum.
Pennebaker, J. W., & Lay, T. C. (2002). Language Use and Personality during Crises: Analyses of Mayor Rudolph Giuliani’s Press Conferences. Journal of Research in
Personality, 36(3), 271–282. https://doi.org/10.1006/jrpe.2002.2349
Pilger, S. (2014, February 6) . Why The Premier League Is the Most Powerful League In the World. Consulted from: https://bleacherreport.com/articles/1948434-why-the-premier-league-is-the-most-powerful-league-in-the-world.
Rabbie, J. M., & Horwitz, M. (1969). Arousal of in-group and out-group bias by a chance win or loss. Journal of Personality and Social Psychology, 13(3), 269–277.
https://doi.org/10.1037/h0028284
Rude, S., Gortner, E.-M., & Pennebaker, J. (2004). Language use of depressed and depression-vulnerable college students. Cognition & Emotion, 18(8), 1121–1133. https://doi.org/10.1080/02699930441000030
Ruder, M. K., & Gill, D. L. (1982). Immediate Effects of Win-Loss on Perceptions of Cohesion in Intramural and Intercollegiate Volleyball Teams. Journal of Sport
Ruscher, J. B. (2001). Prejudiced Communication. New York, The United States of America: Guilford Publications.
Sillars, A., Shellen, W., McIntosh, A., & Pomegranate, M. (1997). Relational characteristics of language: Elaboration and differentiation in marital conversations. Western Journal
of Communication, 61(4), 403–422. https://doi.org/10.1080/10570319709374587
Slodov, A., & Konya, A. (2015). Collective Speech.
Smoll, F. L., Smith, R. E., & Cumming, S. P. (2002). Effects of a Motivational Climate Intervention for Coaches on Changes in Young Athletes’ Achievement Goal Orientations. Journal of Clinical Sport Psychology, 1(1), 23–46.
https://doi.org/10.1123/jcsp.1.1.23
Son, V., Jackson, B., Grove, J. R., & Feltz, D. L. (2011). “I am” versus “we are”: Effects of distinctive variants of self-talk on efficacy beliefs and motor performance. Journal of
Sports Sciences, 29(13), 1417–1424. https://doi.org/10.1080/02640414.2011.593186
Steffens, N. K., & Haslam, S. A. (2013). Power through ‘Us’: Leaders’ Use of We-Referencing Language Predicts Election Victory. PLOS ONE, 8(10), e77952. https://doi.org/10.1371/journal.pone.0077952
Swaen, B. (2017). Conceptual framework: Control Variables. Consulted from: https://www.scribbr.com/dissertation/control-variables/
Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. The social
psychology of intergroup relations, 33(47), 74.
Taylor, J., & Demick, A. (1994). A multidimensional model of momentum in sports. Journal
of Applied Sport Psychology, 6(1), 51–70.
