Race distance - RQ1

A consumer's recognition level is a strong indicator as to whether the message has created a substantial impact or brand awareness (^{Bennett et al., 2006}). Brand awareness is defined as "chhe ability of a consumer to recognise and recall a brand in different situations" (^{Aaker, 1996}, p. 114). In the sponsorship context, brand recognition relates to a consumer's ability to recall past exposure to a sponsor when given a list of several brands as a cue. Brand awareness is a key step in the process of communicating with the consumer and strengthening the brand value. If there is no brand awareness, other communication effects cannot be expected (^{Rogic et al., 2019}). According to ^{Lardinoit and Derbaix (2001}), consumers access information from memory to differentiate between the various sponsors. If the person is able to correctly identify the sponsors, this is indicative that they have taken a considerable interest in the message or event (Bennett et al., 2006).

Consumer awareness of sponsorship is an initial objective, and the impact of that awareness is critical to ultimately determining the effectiveness of a sponsorship deal (^{Biscaia & Rocha, 2018}). The rationale for this assumption is that if brand recognition is not achieved, sponsors will face more difficulties in achieving other subsequent goals, such as favourable attitudes towards sponsors and purchase intentions in relation to their products. It is therefore commonly accepted that increased awareness represents an initial step, or pre-requisite, for a sponsor to obtain a set of potential subsequent benefits.

Although not always from a participation-based sports perspective (^{Eddy & Cork, 2019}), a significant amount of research has focused on the examination of sponsorship awareness (e.g., ^{Biscaia et al., 2013}; Biscaia et al., 2014; Biscaia & Rocha, 2018; ^{Hickman, 2015}; ^{Rogic et al., 2019}; ^{Silva, 2016}; ^{Walsh et al., 2008}; ^{Zaharia et al., 2016}), including the case of race athletics (^{Beaton et al., 2011}; ^{Eagleman & Krohn, 2012}; Eddy & Cork, 2019; ^{Lough et al., 2014}), but none has evaluated the relashionship between the sponsor's brand recognition rate and the distance of the race.

Previous research shows that marathon runners recognised official sponsors that included Zappos.com and others with relevant recognition rates of 97.4%, 73.6% and 80.6% (^{Lough et al., 2014}). At the Fayetteville Race Series event, sponsor recognition rates ranged from 82.5% to 5%, being higher for the real sponsors (^{Eddy & Cork, 2019}). However, ^{Biscaia and Rocha (2018}) did not find completely convergent results, since they found that a greater number of respondents indicated the wrong brands when asked to indicate the cell phone operator and the yogurt brand sponsoring the Rio 2016 Olympic Games.

Based on previous literature and remaining gaps, the first set of research questions (RQ1a, RQ1b) was framed around the ability of runners to correctly recognise the brand of race sponsors, depending on the distance of the race, as shown in Figure 1:

Figure 1.  Research Questions 

RQ1a: Do runners who completed the 10 km race have a higher recognition rate of race sponsors, than runners who completed the 5 km race?

RQ1b: Among runners who correctly identified race sponsors, is the proportion of runners who completed the 10 km race greater than that of runners who completed the 5 km race?

Running involvement - RQ2

Activity involvement is an expression of an individual's interest in a specific sports activity and has been defined as "an unobservable state of motivation, arousal or interest towards a recreational activity or associated product" (^{Havitz & Dimanche, 1997}, p. 246). Involvement is an important factor because of its potential influence on people's attitudes and behaviour in relation to a given product or activity (^{Ridinger et al., 2012}), including the processing of sponsorship information in consumers' minds (^{Dos Santos et al., 2016}). The logic of this relationship seems to lie in a sponsor/sponsored-activity relationship. There is a positive emotional orientation towards the sponsor who bestows benefit on the consumer's favoured activity, in our case, the practice of running (^{Meenaghan, 2001}). This triangular relationship between runner, race (activity) and sponsor brand is shown in Figure 2.

Figure 2.  Hypothesis: Running Involvement and Brand Recognition 

Sports involvement has been shown to have a direct influence on sponsor awareness. ^{Lascu et al. (1995}), for instance, found that individuals (spectators) who were rated higher in sports involvement were more likely to correctly identify the sponsors of a golfing event than those rated lower in sports involvement. Similarly, ^{Levin et al. (2001}) found that respondents with higher levels of sports involvement were better able to recognise and recall sponsor brands displayed as on-car logos at a National Association for Stock Car Auto Racing (NASCAR) event. However, different contexts of basketball, NASCAR, football, and tennis all showed divergent results (^{Bachleda et al., 2016}; ^{Lardinoit & Derbaix, 2001}; ^{Mao et al., 2013}; ^{Walsh et al., 2008}), which were obtained in such a way that the involvement had no effect on awareness. However, it is important to note that the results were found only among sports spectators. In the context of grassroots sports events, athletes, spectators, and volunteers with a higher level of interest in the event, had higher recall rates than those indicating a neutral interest in the same event (^{Miloch & Lambrecht, 2006}). There are only two studies in the context of athletics races: ^{Beaton et al. (2011}) verified that an involvement with the activity in marathon, half-marathon and 5 km runners acted positively on sponsor recognition; and ^{Eagleman and Krohn (2012}), who stated that the identification with running had effects on the sponsor recognition.

Thus, even if it is a little inconsistent, it seems likely that brand recognition is associated with running involvement. Consequently, the second set of research questions was raised:

RQ2a: Do runners with a higher level of running involvement have a higher recognition rate of race sponsors than runners with a lower level of running involvement?

RQ2b: Among runners who correctly identified race sponsors, is the proportion of runners with the higher levels of running involvement higher than the proportion of runners with the lower levels of running involvement?

Loyalty to running - RQ3

The participants' idea of loyalty refers to a process of psychological connection to the sporting object that results in positive, consistent and long-lasting attitudes, as well as associated behaviours (^{Funk & James, 2001}). In other words, behavioural loyalty places the emphasis on observable behaviour recorded in the past, in terms of habits in sporting activity, in which the following were considered: the frequency of running performed during a given period of time and the distance/duration of the runs carried out (^{Downward & Rasciute, 2011}).

The results found are not consistent, ^{Lee et al. (2011}), in the context of recreational golf, found that golfers with a high frequency of golf participation suggested that they have higher sponsorship awareness when compared to other groups (low and mid frequency). This finding is supported by ^{Lascu et al. (1995}), also in the context of golf, although only in relation to spectators. There are, however, other divergent results. ^{Silva (2016}), in the context of judo athletes, did not observe significant differences between the level of practice of judo and the level of recognition of sponsors; it was also found, in the context of NASCAR, that there were no significant effects of the volume of sports consumption on recognition rates (^{Walsh et al., 2008}).

Thus, despite divergent results, it seems likely that brand recognition is associated with loyalty to running. Consequently, the third set of research questions was raised:

RQ3a: Do runners with a higher level of loyalty to running have a higher recognition rate of race sponsors than runners with a lower level of loyalty to running?

RQ3b: Among runners who correctly identified race sponsors, is the proportion of runners with the highest level of loyalty to running higher than that of runners with the lowest level of loyalty?

Data collection

A cross-sectional self-administered survey was conducted. It covered the "Scalabis Night Race" (https://scalabisnightrace.pt/) which has been going for over 12 years. The race has two distances: 10 km and 5 km. The sponsors of the race were four national and local auto brands, petrol supplier, restaurants, a bakery, and a clothing retail outlet, all well known in the region.

Sample and procedures

The individuals were selected from the runners of the Scalabis Night Race, which constituted a sample for convenience. The selection of this specific event was made according to the following criteria: (1) it can be considered as both a leisure and competitive event, but there is no prize money; (2) the event had several co-sponsors, which facilitated the study methodology and the measurement of sponsors' brand recognition; and (3) the event has a good participation rate, with more than 4,000 runners, which is needed to obtain a good sample.

The survey was conducted in May 2018 after the race was completed. An e-mail containing a link to an online survey was sent to all the participants in the race and provided an email address on the entry form.

In total, 3,305 questionnaires were sent, of which 874 were completed and returned. Due to some missing data, 138 individuals were excluded from the analysis. After sorting the data, a total of 736 replies were considered usable surveys, representing a response rate of 18%. According to ^{Bartlett et al. (2001}) we can be 98% confident that estimates from this sample fall within 4% of the corresponding actual population of the race.

Thus, the sample was very representative of the event population. The survey software allowed just one response to be recorded from each IP (Internet Protocol) address, preventing participants from taking the survey multiple times. The ethical criteria adopted were approved by the research centre in which the author is affiliated and respected the provisions of the Declaration of Helsinki. The questionnaires were applied after obtaining informed consent. The study was of the empirical type and used an associative strategy in a comparative process (^{Ato et al., 2013}). The methodology was similar to the one employed by ^{Kim et al. (2010}).

In the current survey, 52.9% of the respondents were men and 47.1% were women. The runners' ages ranged between 18 and 68 years old (M = 41.22, SD = 9.29), with respondents being in the categories of 45-54 years old (27.2%), 35-44 years old (42.0%), 25- 34 years (19.0%), 55-64 years (7.8%), 18-24 years (3.4%) and over 65 years (0.7%). Most respondents were married (46.4%), followed by singles (28.2%), in de facto union (16.2%), divorced (8.3%) and widowed (0.9%). More than half of the runners (63.3%) lived outside the city where the race took place (more than 20 km), designated "national", while 36.7% came from the city of Santarém and less than 20 km, and were designated "local". In the 5 km race, 57.9% of runners were local, while in the 10 km race, 70.3% were national, see Table 1.

Table 1.  Race distance and residence 

The predominant educational qualifications were the level of bachelor's degree (41.6%), secondary education (11th/12th year) (35.6%), master/doctorate (15.7%) and basic education (up to 9th year) completed (7.1 %). The most commonly raced distance by the majority of the respondents was 10 km, 73.4% (n = 540), with 26.6% (n = 196) completing the 5 km run. Race loyalty measured by the number of complete races in the last year showed that the majority of respondents (45.3%) had completed between one and five races under 10 km, 50.2% had completed between one and five races of 10 km, 30.9% had completed between one and five half-marathon runs and 10.1% between one and five marathons, see Table 2. The average number of total races run by respondents each year was 17.88.

Table 2.  Loyalty to running - number of completed athletic races in the last year 

Measurements

For translating and adapting the instruments from the original language to Portuguese, we followed the recommended methodological procedures (^{Vallerand, 1989}). Participants responded to several questionnaires relating to all theoretical concepts under analysis, see Table 3.

Table 3.  Variable definition and measurement 

Statistical Analysis

Means and standard deviations were calculated for all the studied variables. Reliability analysis using IBM SPSS Statistics version 26 demonstrated that the six-item of running involvement (α = .952, exceeded the 0.70 threshold recommended by ^{Nunnally (1978}) and is, therefore, a reliable measure of running involvement. Levels are considered to be a probability of error of .05 in all analyzes.

A series of t-tests were run using 1) two groups of runners who completed the 5 and 10 km run; 2) two groups of runners with levels of running involvement, and; 3) two groups of runners with levels of race loyalty as the independent variable, and the brand recognition sponsor as dependent variable. The assumptions were evaluated, respectively with the Kolmogorov-Smirnov test with Lillefors correction (KS (196) RD5km = .254, p < .000; KS (540) RD10km = .211, p < .000) and with the test Levene test (F (1,734) = 13,792, p < .000) (Maroco 2010, p. 152); (KS (169) RILow = .247, p < .000; (KS (492) RIHigh = .219; p < .000) and with the Levene test (F (1,659) = 6,725, p < .000); Lillefors (KS (544) LOYLow = .221, p < .000; KS (32) LOYHigh = .215, p < .000) and with the Levene test (F (1,574) = .947, p = .331). It is observed that the distribution of variables in the two groups is not normal and the population variances estimated from the two samples are not homogeneous, except for runners with levels of loyalty to the lower and upper running, whose estimated population variances were considered to be homogeneous.

However, the violation of these assumptions has no serious consequences, because the groups or samples have a significant dimension (n = 196, 540, 169, 492), much greater than 40 (^{Myers & Well, 2003}). The skewness (sk) values for the items used in the running involvement (RI) (centrality and attraction) varied from |Sk| = .026 to |Sk| = .583 and the kurtosis (ku) values varied from |Ku| = .651 to |Ku| = .777. For the variable loyalty to running (LOY) they were |Sk| = 1.790 and | Ku | = 3.180 and for the variable brand recognition (BR) they were |Sk| = .753 and |Ku| = .171. According to (^{Kline, 2005}) these values do not represent problems of non-normality. The independence of the runners' residence - local vs national - was evaluated in relation to the distance of the run, the level of running involvement and the level of loyalty to running. It was possible to verify that the distribution of runners by the type of race is dependent not only on residence (χ2 (1) = 39,930, p < .000), but also on the level of running involvement (χ2 (1) = 32,053, p < .000) as well as the level of loyalty to running (χ2 (1) = 8,688, p = .003). Having tested the existence of significant differences between the distance of the run, the level of running involvement and the level of loyalty to running, the existence of statistically significant differences was verified, and in the 5 km run, the level of running involvement was found to be significantly higher (χ2 (1) = 119,219, p < .000), as was the level of loyalty to running (χ2 (1) = 8,941, p = .003).

Recognition rates

A total of 10 brands were presented to the runners, of which only five were true sponsors of the race, and the average brand recognition rate was 76.5% (total number of brand sponsors correctly identified); the average brand recognition error rate was 46.1% (total number of brand sponsors incorrectly identified). The recognition rate varied between 94.7% and 58.7%. All five current sponsor brands were recognised by 37.3% of runners, and more than half of brand sponsors (three out of five) were recognised by 70.8% of runners, with 1.1% of runners not recognising any of the sponsors.

The product categories Petrol (Gas) and Auto, both international brands, were the false sponsors with the highest brand recognition rates, being just 13.0% and 3.9%, respectively, of the true sponsors, see Table 4. The two sponsoring brands with the highest levels of brand recognition were WShopping (94.7%) and Taberna Quinzena (81.3%), both local brands.

Table 4.  Sponsorship recognition rates of the respondents 

The two sponsoring brands with the lowest levels of brand recognition were Repsol (67.3%) and Volkswagen (58.7%), both international brands. True sponsors achieved higher recognition rates than fake sponsors, see Table 4.

Race distance - RQ1

The first set of research questions considered race distance (5 km and 10 km) and its relationship to sponsorship brand recognition. RQ1a aimed to answer the following question: Do runners who have completed the 10 km race have a higher recognition rate of race sponsors than runners who have completed the 5 km race? Considering the 10 sponsor brands presented, the runners who performed the 5 km race obtained an average brand sponsor recognition of 8.06 (M = 8.06, SD = 1.98), while the runners in the 10 km race remembered the sponsors with an average of 7.50 (M = 7.50, SD = 2.37). In the first set of comparisons of the t-tests, it was indicated that the differences observed between the levels of brand sponsor recognition in the two groups of runners were statistically significant (t = 3,212, p = .001), in such a way that the group of runners from the 5 km race featured higher levels of brand recognition than the group of runners who completed the 10 km race, see Table 5.

Table 5.  Sponsorship recognition by distance race, running involvement level and loyalty to running 

RQ1b was set to answer the following: Among runners who correctly identified race sponsors, is the proportion of runners who completed the 10 km run greater than that of runners who completed the 5 km run? The inferential statistical analysis allows us to state that the rate of brand recognition is independent of distance (i.e., it is identical) of the run performed (χ2 (1) = 2,283, p = .131), in such a way that among the runners who identified correctly the race sponsors there were no significant differences between the 5 and the 10 km distance runners, see Table 5 and Figure 3.

Figure 3.  Results of synthesis of research questions 

Running involvement - RQ2

The second set of research questions was related to the evaluation of sponsor brand recognition according to the level of running involvement of runners. Did RQ2a establish that runners with a higher level of running involvement have a higher recognition rate for race sponsors than runners with a lower level of running involvement? Runners with lower levels of running involvement obtained an average brand sponsor recognition rate of 8.02 (M = 8.02, SD = 2.14), while runners with higher levels of running involvement remembered sponsors with an average of 7.57 (M = 7.57, SD = 2.33).

The t-tests indicated that the differences observed between the levels of sponsor brand recognition in the two groups of runners are statistically significant (t = 2,399, p = .020), in such a way that the group of runners with lower levels of running involvement showed higher levels of brand recognition than the group of runners with higher levels of running involvement, see Table 5.

RQ2b aimed to answer the question: among runners who correctly identified race sponsors, is the proportion of runners with the highest level of running involvement (highly involved with the race) greater than that of runners with the lowest level of running involvement? Inferential statistical analysis allows us to state that the rate of brand recognition is independent of the level of running involvement of runners (i.e., it is identical) (χ2 (1) = 1,800, p = .180), so that among runners who correctly identified the sponsors of the race there are no significant differences between the runners of low and high levels, see Table 5 and Figure 3.

Loyalty to running - RQ3

The third set of research questions regarded the evaluation of sponsor brand recognition according to the level of loyalty towards running demonstrated by the runners. RQ3a intended to answer the question: do runners with a higher level of loyalty to running have a recognition rate of race sponsors than runners with a lower level of loyalty to running?

Runners with lower levels of loyalty to running obtained an average sponsor brand recognition of 7.68 (M = 7.68, SD = 2.32) while runners with higher levels of loyalty to running remembered sponsors with an average of 7.50 (M = 7.50, SD = 1.96). The t-tests indicated that the differences observed between the levels of sponsor brand recognition in the two groups of runners are not statistically significant (t = .438, p = .662), in such a way that the group of runners with lower levels of loyalty to running did not show higher levels of brand recognition than the group of runners with higher levels of loyalty to running, see Table 5.

RQ3b was set to answer the following: among runners who correctly identified race sponsors, is the proportion of runners with the highest level of loyalty to running higher than that of runners with the lowest level of loyalty to running? Inferential statistical

analysis allows us to state that the rate of brand recognition is independent of the level of running involvement of runners (i.e., it is identical) (χ2 (1) = 1,271, p = .260), in such a way that among runners who correctly identified the sponsors of the race, there are no significant differences between the runners of low and high levels, see Table 5 and Figure 3

Recognition rates

The aim of this research was to determine the existence of differences in the sponsors' brand recognition rate in runners of distances of 5 and 10 km, in runners with both low and high levels of running involvement, and in runners with low and high levels of loyalty to running. Six distinct research questions were tested.

The recognition rate of current sponsors varied between 94.7% and 58.7%, which can be considered in line with 10 other investigations, of which three were conducted in the area of athletics races. Marathon runners recognised the official sponsors including Zappos.com, GU, and Brooks at rates of 97.4%, 73.6%, and 80.6%, respectively (^{Lough et al., 2014}). In a Fayetteville Race Series, the sponsor recognition rates ranged from 82.5% to 5%, being higher for true sponsors (^{Eddy & Cork, 2019}). However, as has been previously found, a relatively low rate was shown by a web panel of participants; a total of 38% of the final sample recognised Sony as a sponsor of the 2010 FIFA World Cup (^{Mazodier & Quester, 2014}); in the context of sports teams, as many as 90.6% of such respondents managed to accurately name companies sponsoring sports teams, while 72.7% precisely named sponsors of individual athletes (^{Rogic et al., 2019}). In addition, for nine sponsors of the NFL, overall sponsor recognition rates ranged from 80.2%, 74.0%, 30.1% and 37.7% (^{Hickman, 2015}). The sponsor recognition rate varied between 45% and 36% (^{Milochk & Lambrecht, 2006}). Lastly, as would be expected, the recognition (aided-recall) rates showed that official sponsors (true) received higher recognition rates when compared to non-sponsors (fake); these results agree with those obtained by Eddy and Cork (2019) and Lough et al. (2014).

The average rate of brand recognition in the present study was 76.5% (total number of brand sponsors correctly identified), whilst the average rate of recognition of three sponsor brands of judo in Portugal was between 51.7% and 27.2% of athletes who correctly recognised the three sponsoring brands (^{Silva, 2016}). The average brand recognition error rate was 46.1% (total number of brand sponsors incorrectly identified). More than half of the runners (54.8%) incorrectly mentioned "Peugeot" as a sponsor, a similar result to the one obtained by ^{Biscaia et al. (2014}) who verified that 51.2% mentioned incorrectly that a non-sponsor bank was the sponsor of the team.

All five current sponsor brands were recognised by 37.3% of the runners; more than half of the brand sponsors (three out of five) were recognised by 70.8% of runners and 96.7% were able to recognise at least one sponsor. In a similar piece of research, 74.8% of runners recognised at least one sponsor; sponsor recognition for each of the eight series sponsors varied widely (^{Eagleman & Krohn, 2012}), as well as in State Games, a grassroots sports event, where 86% of subjects correctly recognised at least one sponsor with 35% recognising five or more sponsors (^{Miloch & Lambrecht, 2006}). Lastly, nearly all the participants (98.3%) recognised at least one sponsor correctly, and 97.1% correctly recognised seven or more sponsoring brands included in the list (^{Biscaia et al., 2014}).

Relevant results expressed by the high brand recognition rate were obtained by the local brands (WShopping, Taberna Quinzena, Bijou - 94.7%, 81.3% and 80.6%, respectively), despite the fact that most of the runners had national origin (residing more than 20 km from the city). By contrast, international brands (Repsol, Volkswagen) had worse sponsor recognition rate results, both amongst local and national runners. These results show the importance to sponsoring athletic races of promoting local brand awareness linked to the sporting event. To reinforce the previous statement, it is relevant that at the Rio Olympic Games global event, for example, the recognition rate ranged from 28% to 54% in favour of local sponsors (^{Biscaia & Rocha, 2018}).

The recognition rate of sponsors is influenced by many factors, namely brand familiarity and pre-existing brand awareness. As verified in previous research by ^{McAlister et al. (2012}), it has been demonstrated that effective leveraging and activation should result in increased sponsorship awareness, particularly by those participants that are highly identified with the activity. In the absence of pre-event measures, these results should be taken with a degree of caution, since, as has been recognised by researchers in a variety of contexts, brand awareness of sponsors is highly variable, which correspondingly impacts the reliability of measuring a successful investment (^{Miloch & Lambrecht, 2006}).

The results obtained allow us to demonstrate brand recognition as an expression of awareness which constitutes an asset that values brands, in terms of models of consumer perspectives: "Customer-based brand equity" (^{Keller, 2008}), and "Assets categories of brand equity" (^{Aaker, 2005}).

Race distance

The first set of research questions in this study considered race distance (5 km and 10 km) and its relationship to sponsorship brand recognition. RQ1a was based on the idea that runners who completed the 10 km race have a higher rate of recognition of race sponsors than those runners who completed the 5 km race. The result obtained did not allow an affirmative answer to RQ, as it was actually the runners who completed the 5 km race that showed recognition rates of the top sponsors (M = 8.06) while the participants in the 10 km race obtained a lower value (M = 7.50). It is important to note, however, that these surprising differences are statistically significant. A potential explanation for this result may lie in two reasons: the first reason being associated with the 5 km race, which is a race of shorter distance and less physical effort; and the second reason being associated with the likelihood of runners being more focused on objectives of recreation and social interaction, rather than on competitive objectives oriented towards obtaining better scores and personal records. These reasons may allow runners to be more willing to process information from the contextual context of the race (sponsors' brand exposure, sponsors' billboards and flags throughout the race) and this leads to higher brand recognition rates. This phenomenon has been explored and expresses the mere exposure effect. The mere exposure effect occurs when repeated or single exposure to a stimulus, even in the absence of awareness, results in the formation of a positive affective reaction to the stimulus (^{Zajonc, 1968}). The results of decades of research have demonstrated the robustness of the mere phenomenon of exposure with a variety of stimulus domains and levels of awareness in participants (^{Dos Santos, 2016}). Consistent with these ideas, it is generally accepted that repeated exposure is important at increasing the salience of a given brand in consumers' minds (^{Beaton et al., 2011}).

RQ1b established that among the runners who correctly identified the race sponsors, the proportion of runners who completed the 10 km race is greater than that of the runners who completed the 5 km race. The results confirm that the correct identification of all sponsor brands is independent of the distance of the race. It has been proven that, among runners who correctly identified sponsor brands, the existing proportion of runners who completed the 5 km and 10 km run is statistically identical. So, it seems evident that the factors that will have contributed to this totally correct recognition of the sponsors (and not the overall rate that considered the sponsors to be false and true), had effects both on the runners of the 5 km race, as well as on the runners of the 10 km race. Previous literature suggests that consumers' ability to identify a sponsor increases as a function of duration to exposure (^{Biscaia et al., 2014}; ^{Walliser, 2003}), and runners are exposed to sponsors of athletics races on repeated occasions throughout the course of the race in a dynamic context that draws attention to the practice of running and ways to obtain the best running times. From a theoretical point of view this result may seem divergent, as it was in the 10 km run that there were higher levels of running involvement (see Methods), and involvement is an important factor due to its potential influence on people's attitudes and behaviour in relation to a product or activity (^{Ridinger et al., 2012}), including the processing of sponsorship information in consumers' minds (^{Dos Santos et al., 2016}). The most probable explanation for the results seems to come from three key reasons: the first reason being the fact that the runners with local residence are more aware of the brands of local sponsors; the second reason being that most of these runners completed the 5 km run (see Table 1); and the third reason being related to the fact that they showed a higher level of loyalty to running (see Methods); these, therfore are factors that in terms of the mere exposure effect (Dos Santos, 2016) may have contributed to high rates of brand recognition of sponsors.

Running involvement

RQ2a aimed to answer the following question: Do runners with a higher level of running involvement have a better recognition rate of race sponsors than runners with a lower level of running involvement? The results obtained showed the opposite; runners with a lower level of running involvement showed a better performance in brand recognition (M = 8.02) than runners with a higher level of running involvement (M = 7.57), and these surprising differences are statistically significant. Therefore, the answer to RQ2a is negative. In addition, it was not possible to confirm RQ2b which established that among runners who correctly identified race sponsors, the proportion of runners with the highest levels of running involvement is higher than that of runners with the lowest levels of running involvement. In fact, no significant differences were found between the group of runners with the highest levels of involvement and the group with the lowest levels in terms of the correct recognition of all sponsor brands.

From the empirical point of view, this result differs from the results found in the context of running. ^{Beaton et al. (2011}) in the Miami Marathon, half-marathon and 5 km found a direct positive effect of involvement on sponsor recognition. Also, ^{Eagleman and Krohn (2012}), in the US Road Race "chhe Mag 7 Race Series", also found positive effects between identification with the running and sponsor recognition. However, it confirms other results from different contexts: basketball, NASCAR, football and tennis (^{Bachleda et al., 2016}; ^{Lardinoit & Derbaix, 2001}; ^{Mao et al., 2013}; ^{Walsh et al., 2008}). Possible explanations for these results can be given as follows: (1) the specific and distinct nature of running and runners, which differs from that of spectators and television viewers, who passively watch sporting events, and are eventually more available than runners to process sponsor stimuli, and consequently have higher brand recognition rates; (2) the majority (three out of five) of the sponsoring brands of the race are local brands and, according to the tests performed (see Methods), there are significant differences in the level of running involvement between local and national runners, a factor that may explain that runners of local residency, where the levels of running involvement are higher, show higher rates of brand recognition, and (3) perhaps the sponsoring brands did not carry out, on the day of the race and for all runners, enough brand activation actions to generate greater visibility, exposure and relationship, which in terms of repetitive exposure had no effects on awareness levels (^{Lee et al., 2011}). From a theoretical point of view, a key construct in the Elaboration Likelihood Model of persuasion (ELM) (^{Petty & Cacioppo, 1986}) sets out an elaboration likelihood continuum on the central merits of a question or position, in this case, the sponsor brand. However, it appears that sponsor communications (advertisements and posters) may not be a central issue for runners with a higher level of involvement. Perhaps running itself is the centre of attention for runners of 10 km, a fact that may justify low levels of brand recognition.

Loyalty to running

Runners' level of loyalty towards running did not affect the rate of brand recognition. Runners with higher levels of loyalty towards running, who completed the highest number of 10 km runs, half-marathons and marathons in the last year, did not achieve a statistically superior performance of sponsor brand recognition than runners with lower levels of loyalty towards running. The idea formulated in RQ3a, therefore, was not confirmed, meaning that runners with a higher level of loyalty to running do not have a higher recognition rate for race sponsors than runners with a lower level of loyalty to running. Regarding the runners who have correctly identified the race sponsors, there were also no statistically significant differences between the groups of runners with high and low levels of loyalty towards running. It was thus not possible to confirm the RQ3b.

The reasons for such results seem to focus on the distribution of the sample. Most of the sponsoring brands of the race (three out of five) were local brands and two brands were international (Repsol and Volkswagen). According to the tests performed (see Methods), there were significant differences in the level of loyalty to running between runners with local and national residency, an element that may explain that local runners, in which loyalty to running levels are lower, show higher brand recognition rates, especially for local brands, a factor that may have led to the absence of significant differences between the two. Although, repeated exposure to a stimulus (sponsors brands), even in the absence of awareness, results in the formation of a positive affective reaction to the stimulus (^{Zajonc, 1968}) and levels of awareness, it seems evident that the results found are not conclusive. The results obtained agree more with previous investigations into the sponsorship of participation-based sports as opposed to spectator-based sports (^{Eddy & Cork, 2019}): significant differences between the level of practice of judo and the level of recognition of sponsors was not observed by ^{Silva (2016}) in the context of judo athletes; further, in the context of NASCAR, there were no significant effects of the volume of sports consumption on recognition rates (^{Walsh et al., 2008}). However, they differ from those obtained by ^{Lascu et al. (1995}) and ^{Lee et al. (2011}) in the context of golf, in relation to spectator-based and participation-based individuals.

Nevertheless, as highlighted by ^{Kim et al. (2015}) in the extensive literature review, there are complex and inter-connected relationships between the factors influencing sponsorship effectiveness. The different results can be attributed to the widespread diversity of research contexts, such as: (1) for-profit organisations versus not-for-profit organisations; (4) fictitious versus real brands; (3) student versus non-student populations and (4) sponsor's market position.

Implications for race organisers

Firstly, the current study includes information as to how race distance and local brands are related to measures that can be used by race organisers to attract or retain sponsors. Second, from a runner involvement perspective, it seems that race organisers would be wise to focus on runners who are more involved and who participate in longer distance races, since it was this category of runners that showed the lower results. It is suggested that, in coordination with the sponsoring companies, activation activities are planned and executed which may include the offer of gifts of sponsor brands in order to increase brand recognition. This will consequently lead to the formation of favourable attitudes towards the sponsor and increase purchasing behaviours of sponsors' products. Third, spaces next to the start of the race should be created with the objective of developing activities - an incentive programme - to raise the level of running involvement among runners with lower levels of involvement (e.g., autograph sessions with famous runners and the team sponsor). In addition, a presence on the internet and social networks should be maintained in order to facilitate and increase interactions between runners and sponsors, particularly the international brands that obtained the worst results. This information is also relevant in legitimising relationships with sponsors.

Implications for sponsors

The first suggestion is associated with the lack of capacity to create a stronger awareness amongst the most involved runners which not only highlights a missed opportunity, but also signals an unquestionable need for sponsoring companies to focus their leverage and activation efforts on a strategy that maximises runners' interest and involvement. Secondly, a focus for attention should be the segment of runners with the lowest levels of involvement with the activity, which may involve offering various activation and leverage activities by conducting promotional activities at the race site, such as free trials, sweepstakes, or a tent to promote hospitality, in which runners can relax and interact with each other. Thirdly, from the sponsors' perspective, it makes little sense to spend money and effort if the sponsorship agreement does not contribute to the creation of brand value. Thus sponsors should increase, especially in national and international running events, the degree of relationships between runners, and race organisers should favour long-term businesses that help to add historical brand legitimacy and reputation (^{Biscaia et al., 2014}). Finally, given that runners' attentions will most likely be focused on other runners and the course, sponsors' logos should be placed on runners' shirts and in prominent places for runners to see along the course.

Limitations and future research

This research shows some limitations that are worth considering since they provide some direction for future research. First, this study was limited to one athletics race and its associated sponsors, and thus, the results may lack applicability to other sports settings. Further, the adaptation of the involvement measurement scale from ^{Alexandris et al. (2007}) and ^{Ridinger et al. (2012}) may have some limitations. Second, additional samples of runners from different distance races should be collected in future research to clarify further our understanding of sponsorship awareness in athletics races. Data was collected through a mail survey and this may have influenced the sample composition (e.g., local runners). Like the vast majority of investigations (^{Cornwell & Kwon, 2020}), the present study did not include measuring the brand's previous exposure, and it may therefore overestimate the contribution of sponsorship to marketing objectives. Previous studies suggest that long-term deals between sponsors and teams tend to provide benefits to the sponsoring brand even after the sponsorship has ended (^{Biscaia et al., 2014}). Thus, conducting longitudinal studies at different annual events of the same race may contribute to a better understanding of sponsorship awareness among runners and of how to reinforce the race-sponsor link. Fourth, complementarily, it would be important in future studies to assess participants' previous experiences of sponsoring brands, given that product knowledge has been suggested to be influential in a consumer's information processing of sponsorship messages (^{Roy & Cornwell, 2004}). Fifth, while this study focused on brand awareness, other sponsorship outcomes require further attention. ^{Crompton (2004}) infers that individuals pass through a series of stages from first becoming aware of the brand to finally making a purchase decision. Future studies should examine sponsorship awareness and its relationship to additional outcomes such as attitudes towards the sponsor, purchase intentions (^{Speed & Thompson, 2000}) and actual sales (^{Zaharia et al., 2016}), in order to better clarify the benefits that firms obtain when engaging in sports sponsorship. Finally, previous studies mention that the decision to activate a sponsorship programme can influence a firm's culture and motivate employees (^{Chadwick & Thwaites, 2004}). From a perspective of healthy lifestyles through running, addressing the role of sports sponsorship in promoting a more active lifestyle among sponsor employees could be an interesting topic to examine in future research.