Estimation of Emotional and Evaluating Effects of Sport Sponsorships

Citation:

Flemming Hansen, Jens Halling, and Gitte Bach Lauritsen (2001) ,"Estimation of Emotional and Evaluating Effects of Sport Sponsorships", in E - European Advances in Consumer Research Volume 5, eds. Andrea Groeppel-Klien and Frank-Rudolf Esch, Provo, UT : Association for Consumer Research, Pages: 310-317.

European Advances in Consumer Research Volume 5, 2001      Pages 310-317

ESTIMATION OF EMOTIONAL AND EVALUATING EFFECTS OF SPORT SPONSORSHIPS

Flemming Hansen, Copenhagen Business School, Denmark

Jens Halling, Copenhagen Business School, Denmark

Gitte Bach Lauritsen, Copenhagen Business School, Denmark

INTRODUCTION

During the recent decades event marketing and sponsorships have become more and more viible as a supplement or a substitute to traditional advertising. It is therefore important to keep developing methods and instruments to measure the effects of this still progressing phenomenon. The purpose of this paper is to investigate whether it makes sense measuring emotional and evaluating effects of sponsorships, and if this is the caseBto present a standardised instrument, which can be applied to future research.

Traditionally, effects of sponsoring have been looked upon in terms of an effect hierarchy, amount of exposure to banners etc. Recall of sponsorships and image have been studied generally, effects on the latter having been difficult to identify (Hansen & Scotwin 1994). The perception of sponsorships and related "signals" have rarely been looked upon in the lights of low involvement (Rossiter and Percy 1996) or peripheral information processing (Petty & Caccioppo 1988). It is the purpose with this project to develop a procedure that can do so.

METHODOLOGY

Permission was obtained to approach the new 1999 generation of students at Copenhagen business school, a total of app. 450. Data were collected with a mailed questionnaire and the final analysis was conducted on 150 respondents. It is likely that the non response has resulted in a sample, positively biased in terms of interest in sport and sponsoring. 21 sponsorships were included in the questionnaire, covering 15 sports sponsorships and 6 concerning cultural purposes. Furthermore four not existing sponsorships were included, two in each category, making it possible to do further research on the effects of overclaiming. This however, is not done out in this paper.

MEASUREMENTS

The effect measurements here were inspired by the standardised measures used in advertising testing: The (ELAM) Elaboration Likelihood Advertising Model (Hansen 1997).

In the ELAM test, measurements are made reflecting central as well as peripheral information processing. The measurements are illustrated in figure 1. In the upper part, the central processing is found, in the lower part we find the peripheral processing. In central information processing, as in problem solving, the individual is concerned with evaluation of alternatives and their attributes. When advertising is involved, alternatives and their attributes are evaluated. This may be brand/product/company related information.

In the lower part of Fig. 1, the peripheral, less involved, more emotional, information processing is reflected in measurements, relating to the story in the advertising, to aroused emotions, and to the execution of the advertising.

These measures are applied here to the extent that they are meaningful in testing the effect of sponsorships, and the relevant ones are highlighted in the above figure. The use of such standardised measures has the advantage, that it makes it possible to make comparisons with other test areas, such as advertising and design testing (Kristensen et al 2000). The selected measures in the questionnaire are:

1) Awareness of the sponsorships, simply implying if or if not the respondents has knowledge of the sponsorships existence. (1 in figure 1.)

2) Evaluating effects: Attitude towards sponsors. Here, a standard battery of 12 statements was used, and the respondents were asked to choose three statements that they associated with the sponsors. (4 infigure 1)

3) Evaluating effects: Attitude towards sponsorships. Again, a standard battery was used, and the respondents were asked to choose three statements that they associated with the sponsorships. (7 in figure 1)

4) Emotional responses associated with the sponsorship. Again the respondents were asked to choose three statements from a standard battery of 12 items inspired by the ELAM model. (8 in figure 1)

5) Liking. The respondents were asked to report their "all in all" liking of the sponsorships. (9 in figure 1)

6) Preference with regard to sponsor. Here, the respondents were asked to state, to which extent they would recommend the firms behind the sponsorships or their products, to other people. (5 in figure 1)

7) Linking: This refers to the linkage between the two parties involved in the sponsorship and how obvious the respondents experience this. (6 in figure 1)

8) Buying intention. This was partly measured as straightforward "general purchase intention" by any product from the current sponsor, and partly as the tendencies (more or less willing) to buy any product from the sponsor, after being aware of the existence of the sponsorship. (10 in figure 1)

Some of the measurements do not directly concern the sponsorships but only the sponsor. This could especially be a problem when it comes to firms, which have engaged in sponsoring with the primary purpose to increase their credibility or to enhance their company image, rather than to increase sales. Such firms may have little or no interest in how the sponsorship affects buying intention; still we feel that some sponsors indeed could benefit from these measures.

Development of a reduced measurement instrument

To make it possible to cover a large number of sponsorships in the same data collection a briefer version of the questionnaire than the one used here would be preferable. With all the measurements included in the questionnaire, it becomes quite long, and this of course makes it vulnerable to bias. This is particularly so if telephone interviewing were to be used. Hence, it’s reasonable to investigate, if a reduction of the battery can be made without distorting the basic idea and the results.

The reduction process can be divided into two major steps: first the reduction of the two attitudinal and the emotional response batteries is handled, and secondly it is attempted to reduce the remaining 5 effect measures by investigating whether some of them explain the same variation.

FIGURE 1

THE ELAM RESPONSE MODEL (HANSEN, 1997)

TABLE 1

AWARENESS

Since sports sponsorships have by far the largest representation in the Danish market it is decided to concentrate the analyses in this paper on 12 real sports sponsorships, thus, the initial database comprise 1800 sport sponsorship evaluations. Analysis of, and comparisons with "cultural" sponsorships is an object for future research. In making this choice, the fact that the awareness of these cultural sponsorships is very low compared to the sports sponsorships, is also considered. Awareness is shown in table 1. The 12 real sports sponsorships used in this analysis are marked (1), the 4 real cultural sponsorships are marked (2), and the 4 non existing ones are marked (3). (It is seen that only very few respondents claim to be aware of these).

The last sponsorship, CodanBACopenhagen Open" is omitted to avoid two sponsorships from same sponsor. Furthermore it is chosen to carry out this investigation only for respondents having claimed knowledge of the respective sponsorships, reducing the usable number of evaluations to 1038.

Attitudes towards sponsorships

Attitudes towards the sponsorships were measured along the same dimensions as attitudes towards adverting in general and the answers that emerged from this were submitted to a factor analysis.

Both 2, 3 and 4-factor varimax rotated solutions were investigated and seemed to give meaningful interpretations, but the 3-factor varimax rotated solution was chosen because it made good sense and explained 32,4 % of the variance, and only little more variance were explained by including a forth factor.

The solution is shown in table 2. The first factor can be characterised as "Attention value", loading high on not "seen to often", not "unimaginative", "worth remembering" and "exciting". The second factor is loading on "unnatural relation", "different", not "trustworthy" and not "goodwill making". These statements can be seen as expressing, in a negative sense, the degree of trustworthiness associated with the sponsorships. Finally the third factor loads on the negative associations indicating disgust, unappealing and inappropriate.

TABLE 2

THREE FACTOR VARIMAX ROTATED SOLUTION FOR ATTITUDES TOWARDS SPONSORSHIP

TABLE 3

THREE FACTOR VARIMAX ROTATED SOLUTION FOR ATTITUDES TOWARDS SPONSORS

Since each factor in this representation has high loading on two questions, it is recommended that the report and the reduced questionnaire include only two statements from each of the three dimensions.

Attitudes towards sponsor

Similar analyses were carried out on the data concerning attitudes towards the sponsors. Again several factor solutions were investigated and compared with regard to Eigenvalues, explained variance and theoretical interpretations, and a three factor varimax rotated solution was found to be the most useful. Data are shown in table 3.

The three dimensions found here cover very important dimensions of sponsor images. The first is loading on dynamic, modern and not traditional which easily can be associated with new or "trendy" firms. The second factor contains values that appeal to the "common people" indicating simple, social and not technical, and the third dimension covers the more classic, exclusive values such as conservative, prestigious and aesthetic.

Again it is recommended to continue with two statements from each dimension, this being adequate and it does not distort the original interpretation.

Emotions associated with the sponsorships

Emotional responses play a major role in the peripheral information process. These reveal how the ad is received and they act as mediators between exposure and ad liking and brand evaluations.

Again a large battery was used and the data were submitted to factor analyses. This, not surprisingly, revealed that the variations could be explained in two dimensions, a negative and a positive. The first factor is loading highest on the negative emotions such as anger, fear and grief, while the second factor loads high on positive emotions such as joy, enjoyment and happiness. Here six statements should be sufficient to capture the essence from this battery, and these are recommended for use in the reduced version.

Other effect measures

A slightly different approach to the reduction process is used with the remaining effect measures. These measures are concerned with purchase intentions, linking, liking and overall evaluation of or "preference for" the sponsor. Here, it is investigated if some of these measures explain the same variation, and if this can lead to omitting some of them.

All of these perform well in the sense that they are sensitive and seem meaningful to the respondents, however two obvious measures that may be reduced are the two buying intentions. The measure of general purchase intentions is less likely to relate to sponsoring activity, and also in earlier advertising pretests self rated buying intention after exposure has been the most sensitive and the most useful. Additionally the general buying intention has a very high correlation with the preference measure (0,527) indicating that this to a large extent pick up the same variance as the intent measure. With these considerations in mind, it is not felt important to include the general purchase intention, and it is therefore decided to continue with only one buying intention.

TABLE 4

TWO FACTOR VARIMAX ROTATE SOLUTION FOR EMOTIONS ASSOCIATED WITH SPONSORSHIPS

Thus, the overall sponsor battery comprises:

1) Awareness

2) Emotions associated with sponsorships (6 Statements)

3) Attitudes towards sponsorships (6 Statements)

4) Attitudes towards sponsor (6 Statements)

5) Buying intention after knowledge

6) Liking

7) Linking

8) Preference

In addition to this it is recommended to include a measure of the respondents involvement with the sponsored event (or the events associated with the teams or persons sponsored). For this purpose the following scale is recommended:

1) Participates actively in.

2)

3)

4) Sees frequently on TV

5)

6)

7) Not at all interested in

Total Sponsorship effect

In the following it is attempted to create an overall effectscore for sponsorships. Again this is inspired by the ELAM model, and more directly by the ELAM pretest effectscore, used in measuring the effects of and comparing different ads. The ELAM effectscore is a combined and indexed score containing awareness, liking and (self rated change in) buying intention after seeing the ad, making comparisons between different ads possible. The score is computed as I.

FIGURE

With this score a value larger than one indicates a better than average sponsorship and scores below one do the opposite.

In addition with sponsoring it makes sense to include the score reflecting the extent to which the sponsored event, team or person is seen as linked with the sponsor and the products associated with the sponsor. The respondents self rated perceived linking is, when it comes to sponsorships, very important, since it reflect to what an extent they have comprehended the link between the sponsors and the event, team or person. When this is included the ideal score becomes II.

FIGURE

Unfortunately awareness wasn’t formulated on a 5 point scale in the questionnaire making it impossible to include in the effectscore in the following analyses of the reduced battery. It is however recommended to work with a self rated attention score in future applications of the test, making it possible to compute the effectscore based on formula I or II. In the following the overall effectscore is computed as III.

FIGURE

This approach means that for each measure the data are summarised and the average score indexed, before the final score is calculated as the three effect measure scores multiplied. The result of this is shown below ( Table 5).

This table shows large differences in the way sponsorships are perceived on these 3 effect dimensions, and in the following it is possible to give a comparative evaluation of them.

Before that, a few remarks are in order on the use of the integrated sponsor evaluation system introduced here. Two aspects are important.

First, to evaluate how a specific sponsor program works, it is possible to have something to hold it up against. In the present approach, this is accomplished by analysing several different sponsorships.

Secondly, it is important both to be able to judge the overall effect of the activities, and to understand why different effects emerge.

To judge the overall effect, we look upon the effect index introduced above. With reference to the ELAM model (Figure 1), we can see that this measure include the variable numbers 6, 9 and 10, i.e. the variables in the right hand side of the diagram, in the effect end of the system. The measures used to identify the causes behind the effects, are the emotional responses and the attitudes towards the sponsor and towards the sponsorship.

TABLE 5

EFFECTSCORES

TABLE 6

EFFECTSCORES-EVENTS-PERSONS-TEAMS

To the individual sponsor, concern is with his own scores, relative to those of other sponsors, or relative to those of competitors.

The present data are probably biased in several ways. Young students are more involved with sports, and they have other favourites than other segments of the population. Therefore, to illustrate the system by analysing single sponsors, may be more confusing than informative. Instead we have chosen to show some differences between team, event and individual sponsorships, between "good" and "bad" sponsorships and between the sponsoring of soccer and handball.

Different types of sponsorships compared

The 12 sports sponsorships are divided into three groups after type; sponsorships with persons, with teams and with "events", the latter being put in quotation marks since the soccer and handball league aren’t typical events:

Sponsorships with persons:

SASBThomas Bj°rn

ScanboxBBrian Nielsen

B°rsens NyhedsmagasinBChristian Pless

TV2BCamilla Andersen

V6BCamilla Martin

Sponsorships with teams:

CarlsbergBFCK

MD FoodsBHerrelandsholdet i fodbold

Jolly colaBDamelandsholdet i hsndbold

CodanBBr°ndby

Sponsorships with Events:

Faxe KondiBSuperligaen

ByggekramBHerrehsndboldligaen

SparBVM Hsndbold for kvinder

In table 6 it is seen that in the present sample the team sponsorships get the best effectscore whereas the event and person sponsorships have lower scores. Seen in this light table 7,8 and 9 show how the three categories differ in terms of attitudes towards the sponsorships, emotional responses and attitudes towards the sponsors.

In terms of attitudes towards the sponsorships (table 7), teams do best on the first "trustworthiness" dimension, but also give rise to some negative responses in terms of boring and negative attitudes. On these two dimensions, person and event respectively, do the best.

Regarding emotional responses the teams in general create more emotionsBpositive as well as negative.

When it comes to attitudes towards the sponsor, sponsors of teams are seen as most dynamic and social, but not particularly prestigious. Here, not the least, one must have the specific grouping of the sponsorships and the nature of the sample in mind. The team sponsors may simply be more positively looked upon in the first place.

Still, in the present data, it can be sen that teams come out winning, because of the ability to arouse emotions, to be perceived as trustworthy, and to be the most social.

TABLE 7

ATTITUDES TOWARDS SPONSORSHIPS. AVERAGE PERCENTAGE CHOOSING THE STATEMENTS BELONGING TO THE DIMENSIONS USED.

TABLE 8

EMOTIONS ASSOCIATED WITH SPONSORSHIPS. AVERAGE PERCENTAGE CHOOSING THE STATEMENTS BELONGING TO THE DIMENSIONS USED.

TABLE 9

ATTITUDES TOWARDS SPONSORS. AVERAGE PERCENTAGE CHOOSING THE STATEMENTS BELONGING TO THE DIMENSIONS USED.

Winners vs. losers

Another way of investigating the data is to compare sponsorships that in some way are alike, and hereby determine if there are significant differences in the way they are judged by the respondents. This can determine if the effect measures used is capable of making reasonable and clear distinctions between those sponsorships comprehended positively and those comprehended negatively. This was done by finding a "winner" and a "loser" according to the effect score presented in table 5, in some categories. This was done as follows:

Carlsberg 1,05 vs. Codan 0,88 (Soccerteams)

MD Foods 1,44 vs. Jolly Cola 1,08 (National squads)

Faxe Kondi 1,45 vs. Byggekram 0,71 (Leagues)

SAS 1,36 vs. B°rsens Nyhedsmagasin 0,85 (Individuals)

V6 0,96 vs. Scanbox 0,72 (Individuals).

These parings show that there are large differences between comparable sponsorships and its now interesting to investigate if the differences are reflected on the dimensions used in the earlier analyses. This is done simply by calculating the average scores for the five winners and the five losers separately. Results are shown in table 10,11 and 12.

In table 10 the results are very clear; the winners are more trustworthy, less unimaginative and less unappealing when it comes to attitudes towards the sponsorships.

With regard to emotional responses the results are again evident, in table 11 the winners scores higher on the positive and lower on the negative dimension.

Finally in table 12, attitudes towards the sponsors are investigated. Here the interpretation is more diffuse since the dimensions cannot be termed as positive or negative. Different companies have different values and goals; simple/social can be the wanted image for some, aesthetic/prestigious for others. Also the scores here are without doubt to a large degree explained in terms of the actual companies chosen for this analysis, rather than in terms of sponsoring activities.

Handball vs. Soccer

I Denmark the by far most popular sports, measured by number of active players, press coverage, television viewing etc., are soccer and handball. This also applies in the population analysed here.

Table 13 shows that soccer sponsorships give rise to much higher effect scores than handball. As suggested earlier, the student body include a slight male majority which in any event would favour soccer interests versus handball interests. However, the differences between the two kinds of sports sponsorships are so marked that it completely overrules this potential bias. Additionally the difference rest on marked and meaningful differences in the perception of the two sets of sponsorships (In the database a total of four handball sponsorships and similarly a total of four soccer sponsorships are included).

In terms of attitude towards the sponsorships (table 14) soccer is seen as more trustworthy and less unimaginative and less "seen to often".

Emotionally it generates clearly more positive responses (table 15).

And in terms of evaluation of the sponsor (table 16), soccer sponsors maybe seen as more prestigious, dynamic and social straightforward. The latter of course may be ascribed to differences in the companies choosing to sponsor the different events.

Considering the nature of sample studied here no further attempt should however be made to carry out generalisations with regard to the appropriateness of the two different sport areas for sponsoring.

TABLE 10

ATTITUDES TOWARDS SPONSORSHIPS

TABLE 11

EMOTIONS ASSOCIATED WITH SPONSORSHIPS

TABLE 12

ATTITUDES TOWARDS SPONSORS

TABLE 13

EFFECTSCORES-HANDBALL-SOCCER

TABLE 14

ATTITUDES TOWARDS SPONSORSHIPS

TABLE 15

EMOTIONS ASSOCIATED WITH SPONSORSHIPS

TABLE 16

ATTITUDES TOWARDS SPONSORS

CONCLUSION

Traditionally sponsoring has been evaluated in terms of its ability to generate exposure, recall, and occasionally attitude to or preference changes towards the sponsor. The latter, however, has in most cases have been difficult to establish.

Departing in contemporary theories of communication concerned with low involvement, emotional peripheral ways of coping with information, it has been attempted here to demonstrate, that much more significant effects of sponsoring can be established when looking upon attitudes towards the sponsorship, liking, linking between sponsor and sponsorship, emotional responses etc.

The study has been carried out with a very extensive questionnaire to make sure that all possible details of such evaluations would be uncovered. Using a sample of qualified but biased respondents in terms of younger better educated persons of both sexes it has been established that remarkable differences can found in the way in which different sponsorships are evaluated. These effects by far overrule the more traditional measures of effect relying upon the idea that sponsoring should be studied as a effect hierarchy, and that its significance should be evaluated in the light of its ability to influence attitudes towards the sponsor, purchase intention, preferences and related behaviour.

In the project focus has been only on the effects of sponsorships generated by communication of the sponsorship trough the mass media. For many sponsors a number of other related activities such as inviting guests, providing gifts etc. are of importance also. These effects are left out of sight here. With an increasing acceptance of the role emotional unconscious responses play in advertising (Rasmussen, Damasio, Ambler, Aaker, Richins) it is obvious to expect that the effects of sponsorships may be seen in this light. The present study proves that this can be done meaningfully. To extent the findings and to make the observations more generally applicable, a wider selection of sponsorships and a more representative sample is required. To make such studies feasible a reduction of the questionnaire instrument used here has been part of the reporting of the findings shown. It is the hope of the research team to be able to carry out such more general studies with larger samples in the near future.

REFERENCES

Hansen, F. & Scotwin, L.: "The effect of sponsoring: An experimental study". Research Paper, Copenhagen Business School, Dept. of Marketing 1994

Rossiter, J. & Percy, L.: "Adverting Communications & Promotion Management". McGraw-Hill 1997

Petty, R., Cacioppo, J. & Schumann, D.: "Central and Peripherical Routes to Advertising Effectiveness: The Moderating Role of Involvement". Journal of Consumer Research, Vol. 10. 1983.

Hansen, F.: "Quantifying Creative Contributions: Advertising pretesting¦s new generation" in Proceedings from ESOMAR, Edinburgh 1997

Rasmussen, A.: "What is emotions?" in Contributions to marketing. Gallup Denmark, 1996

Damasio, A.: "Descartes¦ Error". Avons Books, 1995

Ambler, T. & Burne, T.: "The impact of affect on memory of advertising". Journal of advertising research, March/April 1999

Aaker, J. & Willams, P.: "Empathy versus pride: The influence of emotional appeals across Cultures". 1998.

Richins, M.: "Measuring Emotions in the Consumption Experience". Journal of Consumer Research, Vol. 24 1997.

----------------------------------------