Document Sample
					           Schmalenbach Business Review ◆ Vol. 56 ◆ April 2004 ◆ pp. 119 – 138

Reinhard Grohs/Udo Wagner/Sabine Vsetecka*



Recent surveys find that while managers favored issues of media coverage not more than
ten years ago, now they rate sponsor awareness and image transfer from the sponsored
event to the sponsor as the main reasons for engaging in sport C. However, the evalua-
tion of C has not kept up with this change in priorities. Companies seem to be reluctant to
evaluate sponsor awareness even though measurement is straightforward and not very
costly. An important reason might be that previous studies showed unsatisfying effects of
“ambush marketing”. In this form of marketing, other firms make consumers believe, incor-
rectly, that these companies are the actual sponsors of an event. In the case of image
transfer, evaluation seems to be difficult due to a lack of a compelling comprehensive and
testable model.

We examine these obstacles of assessing sponsor awareness and image transfer in sport
sponsorships in two ways. As a means of reducing the danger of ambush marketing, we
analyze what drives correct sponsor identification. Our empirical results indicate that
event-sponsor fit, event involvement, and exposure are the dominant factors predicting
sponsor recall. These factors offer sponsors a basis for successful sponsorship planning
and execution through the selection of an appropriate sponsorship. Second, we propose
and empirically test a model that assesses image transfer in sport sponsorships. We find
support for a basic level of image transfer for all sponsors. However, more detailed
research and interpretation of results suggest that the magnitude of image transfer
depends on two factors, sponsorship leverage and event-sponsor fit.

JEL-Classification: M31.

Keywords: Ambush Marketing; Image Transfer; Sponsor Awareness; Sponsorships;
          Sponsorship Effectiveness.


Until the beginning of the 1990s, managers rated media coverage as the main
objective of companies involved in sport sponsoring 1. Since then, mere exposure
has lost lots of its appeal, primarily because the effect of media coverage on con-
sumers’ attitudes or behaviors towards a certain brand is very weak, if measurable
at all. Sponsorship goals have moved towards matters more directly associated

 * Mag. Reinhard Grohs, o.Univ.-Prof. Dr. Udo Wagner, Mag. Sabine Vsetecka, Marketing Department,
   University of Vienna, Brünner Strasse 72, A-1210 Wien, Austria.
 1 See Abratt et al. (1987); Meerabeau et al. (1991); Scott/Suchard (1992).

sbr 56 (2/2004)                                                                             119
R. Grohs/U. Wagner/S. Vsetecka

with consumer behavior. Recent surveys in Great Britain 2, Canada 3, and Germany 4
show that now boosting brand awareness and company image are the dominant
reasons for firms to get involved in sponsorships.

Even though sponsorship goals have changed, the ways of evaluating sponsorship
success remain the same. When we ask managers how they assess the success of
their sponsorships, measuring media coverage and exposure is still the most
important tool. Studies by Hermanns 5 and Thwaites et al.6 suggest that managers
use this method for approximately 70 percent of sponsorships. Use of other evalu-
ation tools to measure brand awareness due to a sponsorship is far less common,
at a rate of between 20 and 30 percent. Comprehensive image studies are particu-
larly rare except for large sport sponsorships with high sponsorship fees, such as
the soccer World Cup 7 or the Olympic Games 8.

A study by Thjomoe et al.9 asked Norwegian firms directly whether they used
appropriate research methods to determine if they achieved their sponsorship
goals. The authors found that of 76 firms that agreed with “we sponsor to increase
awareness”, only 15,4 percent agreed with the statement “we study brand percep-
tions before and after sponsorship”. Of the 69 companies that agreed on “we
sponsor to improve our image”, only 15,6 percent studied brand perceptions
before and after the sponsorship.

These findings raise a straightforward question: Why does only a small minority of
firms choose adequate measures to determine whether they achieved their spon-
sorship goals? In the case of awareness goals, the reasons are far from obvious,
since the measurement of sponsor awareness is easy, for example, by using both
unaided and aided recall as well as recognition10.

One reason might be that field studies have produced unsatisfactory results for
many sponsors in recent years. Sandler/Shani 11 asked viewers of the 1988 Winter
Olympic Games who they thought were official sponsors, and found that viewers
recalled the official sponsors more often than nonsponsors in only four of seven
product categories. Eleven of the 20 brands most often identified as worldwide
sponsors of the 1998 Winter Olympics were not in fact sponsors 12. The phenome-
non of consumers incorrectly believing that a certain company is an actual spon-
sor of an event is called “ambush marketing” 13. This form of marketing might
deter sponsors from using recall to measure their sponsorship success. However, if
willingness to assess sponsor recall is low because results are unsatisfying, spon-

 2   See   Thwaites (1995).
 3   See   Thwaites et al. (1998).
 4   See   Hermanns (2000).
 5   See   Hermanns (2000).
 6   See   Thwaites et al. (1998).
 7   See   Otker/Hayes (1987).
 8   See   Stipp/Schiavone (1996).
 9   See   Thjomoe et al. (2002).
10   See   Keller (1993).
11   See   Sandler/Shani (1989).
12   See   The Wall Street Journal (1998).
13   See   Sandler/Shani (1989).

120                                                                   sbr 56 (2/2004)
                                                                           Sport Sponsorships

sors should be primarily concerned with determining factors that increase correct
recall of an official sponsor. If such guidelines can be established and successfully
integrated into event sponsorships, then firms might become less reluctant to mea-
sure their success by means of awareness studies.

In this paper we identify drivers of correct and enhanced sponsorship recall (such
as event-sponsor fit and event involvement) and test them empirically for various
industries at a sponsored sport event.

Achievement of image related sponsorship goals is more difficult to evaluate. The
main task is how to ensure that the image change is accounted for by the spon-
sorship, and that it is not due to external factors such as carryover-effects from
other promotional instruments, interdependencies, or new competitors 14. Previous
research suggests a number of approaches 15, but none of these studies was able
to fully comprehend the effects that sponsorship has on the sponsor’s image. We
assess the merits and shortcomings of these approaches. We contribute to the
ongoing discussion by proposing a model that explains post-event sponsor image
as a function of pre-event sponsor image, sponsor awareness, and event image.
To assess the effectiveness of image transfer in sport sponsorships we test our
model for various branches at a sponsored sport event. Finally, we identify condi-
tions (such as industry type of the sponsor or event-sponsor fit) under which an
image transfer is most likely to take place.



Brand awareness is a basic concept behind most brand-related goals that aim at
marketing goods and services, such as positioning or brand image. If awareness is
the main sponsorship goal for a company, the key question at the end of the day
is how well does the recipient (in this study, the event visitor) remember the
sponsoring firm?

Most empirical studies have used tracking measures to evaluate the awareness
engendered by sponsorships 16. These methods are widely used and numerous
examples can be found 17. However, quite a few authors have discovered that con-
sumers get confused about official sponsors, because nonsponsors often try to
build up an association with a particular event, too 18. In recent years it has
become more and more of a challenge to design and communicate sponsorships
in a way that ensures that visitors classify sponsors and nonsponsors correctly, and
that nonsponsors’ ambush marketing is prevented. To accomplish this goal, spon-
sors must understand the factors that determine sponsor recall. In this paper we
theoretically analyze these factors and develop hypotheses that we test empiri-

14   See Deimel (1992); Meenaghan (1991).
15   See, e.g., Crimmins/Horn (1996); Quester/Thompson (2001); Stipp/Schiavone (1996).
16   See Cornwell/Maignan (1998).
17   See, e.g., Crimmins/Horn (1996); Lardinoit/Derbaix (2001).
18   See, e.g., Crimmins/Horn (1996); Sandler/Shani (1989).

sbr 56 (2/2004)                                                                           121
R. Grohs/U. Wagner/S. Vsetecka

cally. Building on the insights gained, creation and execution of successful spon-
sorships with enhanced sponsor awareness can become better understood and
easier to achieve.
Brand Prominence
Because sponsorship messages cannot pass on specific information about a prod-
uct or service, audiences need to have a basic knowledge about the sponsor and
its product category. The consumer can then connect the information conveyed by
the sponsorship to established knowledge and value structures, thus enhancing
sponsor recall 19. According to Glogger 20, sponsor identification is also more accu-
rate if previously seen logos and signs are used. Consequently, sponsor identifica-
tion may involve a brand prominence bias. Johar/Pham 21 show empirically that
consumers use the heuristics of market prominence to infer the sponsor of an
event. We note that such heuristics might also boost prominent ambush marketers
who seek improper identification. Hence, preceding evidence suggests that:
– H.1.1: Brand prominence is positively associated with sponsor awareness.
Event-Sponsor Fit
Scientific literature has used numerous words to describe the fit between a spon-
sor and a sponsored activity, such as synergy 22, similarity 23 or link 24. Generally,
most authors distinguish between a functional fit and an image related fit 25. Func-
tional fit describes the thematic relatedness between a sponsor and an event 26.
Image related fit encompasses the attributes associated with a sponsor and a spon-
sored event 27. Because audiences misidentified nonsponsors as sponsors at vari-
ous activities, researchers became interested in how consumers distinguished
between official sponsors and ambush marketers 28. Results indicate that con-
sumers might invoke heuristics to recall sponsors. All else equal, companies and
brands that seem to be related to an event are more likely to be identified as
actual sponsors 29. Consumers use event-sponsor fit as a source of information
when inferring the identity of event sponsors, particularly if a sponsor is recalled
some time after the event. Therefore, we propose that:
– H.1.2: Perceived event-sponsor fit is positively associated with sponsor aware-

19 See Deimel (1992).
20 See Glogger (1999).
21 See Johar/Pham (1999).
22 See McDonald (1991).
23 See Gwinner (1997).
24 See Otker/Hayes (1987).
25 See, e.g., Gwinner (1997).
26 For example, if Adidas, a manufacturer of sport shoes, sponsors the soccer World Cup, functional
   similarity is high.
27 For example, if Red Bull, a manufacturer of soft drinks that promotes its product as “stimulating
   users for exaggerated action”, sponsors extreme sport competitions like base jumping events,
   image related fit is high.
28 See Johar/Pham (1999); Quester/Farrelly (1998).
29 See Pham/Johar (2001).

122                                                                                 sbr 56 (2/2004)
                                                                             Sport Sponsorships

Event Involvement

Meenaghan 30 examines the concept of fan involvement and its implications for
sponsorships. Focus group interviews show that increased event or fan involve-
ment in a particular sponsored activity evokes a positive emotional orientation
towards the sponsor. Highly involved fans are the most aware of the sponsor’s
investments. Indeed, research at the soccer World Cup 1998 in France showed that
although 55 percent of respondents could spontaneously recall any sponsor, this
result rose to 80 percent for fans who identified themselves as being “very inter-
ested” in the soccer World Cup. High involvement further reduced the number of
incorrect sponsor attributions from 67 to 56 percent 31.

Harvey 32 develops a similar hypothesis but fails to test it empirically. Hansen/
Scotwin 33 present further evidence that more involved sports fans have signifi-
cantly higher recall than do those not involved for one sponsor, but fail to find
this relation for a second sponsor. Based on these findings we hypothesize that:

– H.1.3: Event involvement is positively associated with sponsor awareness.


The standard finding in cognitive learning is that message learning grows with
additional exposures, although at a diminishing rate. Furthermore, the effective-
ness of repetition is higher if the recipient’s involvement in the message content is
low. For sports sponsorships, interest in the sponsor is low but involvement in the
sponsored activity is high. Indeed, many studies find a positive impact for the
duration of a sponsorship on the awareness of the sponsor of the event 34. There-
fore, we posit that:

– H.1.4: Exposure to the sponsored event is positively associated with sponsor

The left half of Figure 1 displays a graphical representation of the comprehensive
conceptual model. It relates the four factors brand prominence, event-sponsor fit,
event involvement, and exposure to sponsor awareness.


Image transfer in sports sponsorships is defined as the transfer of associations
attributed to the sponsored activity to the sponsoring brand 35. The aim is to evoke
positive feelings and attitudes towards the sponsor by closely linking the sponsor

30   See   Meenaghan (2001).
31   See   Meenaghan (2001).
32   See   Harvey (2001).
33   See   Hansen/Scotwin (1995).
34   See   Cornwell/Roy/Steinard (2001); Crimmins/Horn (1996); d’Astous/Bitz (1995).
35   See   Gwinner (1997).

sbr 56 (2/2004)                                                                             123
R. Grohs/U. Wagner/S. Vsetecka

to an event the recipient values highly. In other words, the image of the event
should “rub off” on the sponsors.

Although research on brand (and also event) image is abundant, there is consider-
ably less agreement on its appropriate definition 36. For our purpose, we define
brand (event) image in accordance with Keller as “… perceptions about a brand
(an event) as reflected by the brand (event) associations held in consumer mem-
ory. Brand (event) associations are informational nodes linked to the brand (event)
node in memory and contain the meaning of the brand (event) for consumers”37.
Associations linked to sports events may be health, youth, vitality, fairness, free-
dom or engagement 38. However, different kinds of sports sponsorships might be
related to very different image dimensions. Although competitive sports are typi-
cally connected with success and strain, many team sports stand for ideas of
group spirit and reliance. Sponsors-to-be must embark on sports sponsorships that
convey the image a sponsor wants to communicate.

Evaluating the image transfer is vital for assessing the effectiveness of a sponsor-
ship. Yet so far, only a small number of studies empirically assess the image trans-
fer from a sponsored event to a sponsoring company 39. Except for one 40, all of
these studies are concerned with sports sponsorships. Overall, we can identify two
concepts for assessing image transfer.

• The first procedure measures the image of the sponsor (the image dimensions
  the sponsor considers to be relevant) before and after the sponsored event. The
  change in image is then attributed to the sponsorship. Several researchers use
  this method 41. Results show partial support for successful image transfers in
  sports sponsoring. Crimmins/Horn 42, Otker/Hayes 43 as well as Rajaretnam 44
  find a weak but consistently significant image change. Javalgi et al.45 conclude
  the change in sponsor image depends on the industry. Hansen/Scotwin 46 find
  no statistically significant support for the hypothesis that sponsorships cause an
  image change for the sponsoring brand. Explaining their results, Hansen/
  Scotwin 47 cite the fact that attitudes are stable and that their experimental
  design was not developed for measuring attitude changes.

36   See Keller (1993).
37   Keller (1993), p. 3, ((event) added).
38   See, e.g., Drees (1992).
39   See Crimmins/Horn (1996); Gwinner/Eaton (1999); Hansen/Scotwin (1995); Javalgi et al. (1994);
     Otker/Hayes (1987); Quester/Thompson (2001); Rajaretnam (1994); Stipp/Schiavone (1996).
40   See Quester/Thompson (2001).
41   See Crimmins/Horn (1996); Hansen/Scotwin (1995); Javalgi et al. (1994); Otker/Hayes (1987);
     Quester/Thompson (2001); Rajaretnam (1994).
42   See Crimmins/Horn (1996).
43   See Otker/Hayes (1987).
44   See Rajaretnam (1994).
45   See Javalgi et al. (1994).
46   See Hansen/Scotwin (1995).
47   See Hansen/Scotwin (1995).

124                                                                               sbr 56 (2/2004)
                                                                     Sport Sponsorships

     A notable problem associated with pre- and post-event measurement of spon-
     sor image is that the “before” measure of sponsor image might sensitize respon-
     dents and distort their reaction to the experimental treatment (testing effect 48).
     Furthermore, most of the studies do not use adequate control groups to ensure
     that the image change is indeed accounted for by the sponsorship 49.

• The other method for assessing image transfer does not look at actual changes
  in sponsor image. Instead, it examines the incremental effect event image has
  on sponsor image after the event. Image transfer should be more pronounced if
  the association between event image and sponsor image is high 50. Stipp/Schi-
  avone 51 use this concept in a slightly adjusted version and find that in general,
  attitudes towards Olympic sponsors had a highly significant positive impact on
  the images of specific actual sponsors.

     Apart from Gwinner/Eaton 52 and Stipp/Schiavone 53 we do not find any other
     study that assesses the relation between event and sponsor image empirically,
     even though this procedure is suggested in some conceptual work 54. One rea-
     son and maybe the biggest shortcoming might be the incompleteness of this
     method. The model only proposes event image to affect post-event sponsor
     image. The model ignores the more obvious predictors such as attitudes
     towards the sponsor before the event.

In our study we develop a model for assessing image transfer that captures the
ideas of both concepts explained in the previous paragraphs. We assume that
according to attitude theory, the image of a sponsor is stable over time 55. This
assumption leads to the conclusion that pre-event sponsor image should have a
significant impact on post-event sponsor image. Furthermore, research on classical
conditioning finds a positive effect of awareness on attitudes, referred to as the
mere exposure effect 56. For sponsorships, this effect suggests that sponsors whose
messages create higher recall will have a more positive post-event sponsor image.
The concept of image transfer states that sponsor image after the event is also
determined by event image 57. In other words, event image should have a signifi-
cant incremental contribution in shaping attitudes towards the sponsor after the
event. Hence, post-event sponsor image is a function of pre-event sponsor image,
sponsor awareness, and event image. The right half of Figure 1 shows a graphical
representation of these relations. Formally, we propose:

48   See, e.g., Churchill/Iacobucci (2002), pp. 144 f.
49   For an appropriate study design, see Quester/Thompson (2001).
50   See Gwinner/Eaton (1999).
51   See Stipp/Schiavone (1996).
52   See Gwinner/Eaton (1999).
53   See Stipp/Schiavone (1996).
54   See, e.g., Gwinner (1997).
55   See Kroeber-Riel/Weinberg (2003), p. 179.
56   See Zajonc (1968).
57   See Gwinner (1997).

sbr 56 (2/2004)                                                                     125
R. Grohs/U. Wagner/S. Vsetecka

– H.2.1: Pre-event sponsor image has a positive impact on post-event sponsor

– H.2.2: Sponsor awareness has a positive impact on post-event sponsor image.

– H.2.3: Event image has a positive impact on post-event sponsor image.



We collect data from before and after the Alpine Ski World Championships 2001 in
St. Anton, Austria, from January 28th to February 10th. In the beginning of January,
we asked 200 previously selected individuals from all over Austria to evaluate the
images of six brands from different industries: BMW (automobile), Milka (choco-
late), Telekom Austria (telecommunication), Tag Heuer (watch), Memphis (ciga-
rette and travel agency) and Carlsberg (beer). At the time, the respondents were
not aware that these firms were main sponsors of the World Championships. In
March, the same persons rated the six brands again, now aware of the sponsor-
ships. These people also answered several other questions related to our study.
After eliminating respondents who failed to fill out the second questionnaire, the
final sample consisted of 132 pairs of pre- and post-event responses.


For the first questionnaire, we developed scales for measuring brand image of the
six sponsors. The second questionnaire measured brand image in the same way
again, and also the event image of the Alpine Ski World Championships. We also
measured sponsor recall and drivers determining recall, namely brand promi-
nence, event-sponsor fit, event involvement and exposure 58. We assessed the
leverage of sponsorships independently by collecting data on other communica-
tion activities for these six sponsors that supported their engagement in the World

Sponsor Awareness

We used a common measure of sponsor awareness, unaided recall, as well as
aided recall, with the industry type of the sponsors as a cue (see Appendix A).
This operationalization is suggested in literature 59 and has been used successfully
in sponsorship research 60.

58 The questions and corresponding response categories for all variables used in the study can be
   found in Appendix A.
59 See Keller (1993).
60 See, e.g., Lardinoit/Derbaix (2001).

126                                                                              sbr 56 (2/2004)
                                                                        Sport Sponsorships

Brand Prominence

We assessed brand prominence by asking respondents before the sponsorship
whether or not they were familiar with a particular brand (see Appendix A). Mea-
surement of this variable restricts sample size for further analysis in two respects.
To determine the impact of brand prominence on sponsor recall, it is necessary
that enough respondents are not familiar with the sponsor before the event. Oth-
erwise, there is no way of assessing the effect of high sponsor prominence com-
pared with low sponsor prominence (there would be no low sponsor prominence
under such circumstances). Conversely, to empirically test the relation between
pre- and post-event sponsor images we need a considerable sponsor prominence
before the event, because otherwise, respondents are not able to assess pre-event
sponsor image.

Empirical data showed that more than 90 percent of the respondents knew four of
the six sponsors (BMW, Milka, Memphis, Telekom Austria) before the event, but
the other two sponsors (Carlsberg, Tag Heuer) were familiar to only approxi-
mately 60 percent of the respondents. Therefore, we could test H.1.1 only for
Carlsberg and Tag Heuer. On the other hand, we performed the image transfer
hypothesis testing with a slightly smaller number of respondents 61 for Carlsberg
and Tag Heuer than for the other brands. However, this fact should not have a sig-
nificant impact on the key results.

Event-Sponsor Fit

We measure event-sponsor fit by using a modified three-item, five-point rating
scale proposed by Gwinner/Eaton 62 (see Appendix A). Exploratory factor analyses
for all six sponsors ensured that the three items measure one construct, event-
sponsor fit (one factor explained at least 70 percent of total variance for every
sponsor). Cronbach alphas greater than 0.75 in all cases also support the reliability
of the scale.

Event Involvement

To measure event involvement, respondents rated how interested they were in the
Alpine Ski World Championships 2001 on a five-point scale (see Appendix A). This
operationalization follows prior research by Hansen/Scotwin 63 and Meenaghan 64.


We asked respondents how often they watched television broadcasts from the
Alpine Ski World Championships 2001. We measured exposure in five categories,
ranging from “daily” to “not at all” (see Appendix A).

61   We exclude those respondents who did not know these two brands beforehand.
62   See Gwinner/Eaton (1999).
63   See Hansen/Scotwin (1995).
64   See Meenaghan (2001).

sbr 56 (2/2004)                                                                        127
R. Grohs/U. Wagner/S. Vsetecka

Sponsor and Event Images

To measure sponsor and event images, we developed items based on prior
research on sports image 65 and a brainstorming session with a group of masters
candidate students. Two judges eliminated those items that did not have a mean-
ingful relation either with the World Championships or with any of the six spon-
sors. The respondents rated the remaining 12 items sport, dynamic, health, suc-
cess, Austria, tradition, youth, precision, risk, sociableness, enthusiasm and winter
on five-point scales that reflected the perceived association with the sponsors and
the event (see Appendix A). We measured both event and sponsor images before
and after the event as the aggregate of these 12 image dimensions.


For our data analysis, we first present an overview of general sponsor awareness.
We then empirically assess the effects of the identified drivers on sponsor recall.
Second, we design a model that decomposes post-event sponsor image into four
components: pre-event sponsor image, sponsor awareness, and a composite factor
that accounts for both event image and for sponsor-specific effects in image trans-
fer (such as sponsorship leverage and originating industry).


We find the highest sponsor recall (aided on the top of unaided recall) for Milka,
with 86 percent of the 132 respondents, followed by Telekom Austria (69 percent),
BMW (53 percent), Tag Heuer (36 percent), Memphis (30 percent), and Carlsberg
(24 percent). In a relative sense Carlsberg performs especially well in terms of
unaided recall and Tag Heuer and Memphis score particularly high on the aided
recall measure.

Table 1 presents the awareness results for all six sponsors. Evidence of ambush
marketing is noticeable for Carlsberg and Tag Heuer, but does not seem alarm-
ingly high. With the cue “beer” 15 percent of the respondents incorrectly attrib-

Table 1: Sponsor awareness

 Sponsor                       Unaided Recall          Aided Recall *           Total Recall
 Milka                             62%                     24%                      86%
 Telekom Austria                   44%                     25%                      69%
 BMW                               31%                     22%                      53%
 Tag Heuer                         13%                     23%                      36%
 Memphis                            5%                     25%                      30%
 Carlsberg                         16%                      8%                      24%

* A cue was provided only if the respondent failed to remember the sponsor beforehand.

65 See Drees (1992); Dreyer (1986).

128                                                                                sbr 56 (2/2004)
                                                                                Sport Sponsorships

uted the sponsorship to companies other than Carlsberg, for “watches” the corre-
sponding number is 10 percent. For all other sponsors, misidentification is lower
than 3 percent.

Brand Prominence and Sponsor Awareness

According to Hypothesis H.1.1, people who know the sponsor beforehand should
have a higher sponsor awareness after the event.

We could test this hypothesis empirically only for Carlsberg and Tag Heuer, since
more than 90 percent of respondents were familiar with all other sponsors before
the sponsorship (c.f. section 3.2). For Carlsberg and Tag Heuer we used χ2-tests to
establish effects of brand prominence (sponsor unknown/known) on recall (no
recall/aided recall/unaided recall). We found no significant effects for Carlsberg
(p = 0,34), but we did find evidence in favor of Hypothesis H.1.1 in case of Tag
Heuer (p < 0,01).

Event-Sponsor Fit and Sponsor Awareness

Hypothesis H.1.2 proposes that people who perceive a closer fit between event
and sponsor will remember the sponsor better.

To test this hypothesis, we first carried out a median split of the sample for each
sponsor with one half consisting of respondents who perceived a high fit, the
other one consisting of respondents perceiving a low fit. We then tested our
hypothesis using χ2-tests (low/high fit; no recall/aided recall/unaided recall). We
found a highly significant (p < 0,01) positive relation between perceived fit and
sponsor recall for all six sponsors. Our results support Hypothesis H.1.2.

Event Involvement and Sponsor Awareness

Hypothesis H.1.3 states that people who are highly involved in the Alpine Ski
World Championships will be more likely to remember the sponsors of the event
than will those people whose involvement is low.

We tested this hypothesis by regressing event involvement on sponsor recall (no
recall/aided recall/unaided recall). Results of a multinominal logistic regression
model support our assumption for all sponsors.

Parameter estimates suggest that event involvement has a significant effect on dis-
tinguishing between no recall and unaided recall (p < 0,05 for Memphis, p < 0,01
for all other sponsors). However, we found no evidence that event involvement
significantly affects no recall compared to aided recall unless for Tag Heuer
(p < 0,01). The odds 66 that a viewer will remember a sponsor unaided increase by
between 65 percent (for BMW) and 40 percent (for Telekom Austria) when event
involvement rises by one level. Overall, we find partial support for Hypothesis

66 For interpretation of the so called “odds ratio” see, e.g., Backhaus et al. (2003), pp. 434 f.

sbr 56 (2/2004)                                                                                     129
R. Grohs/U. Wagner/S. Vsetecka

Exposure and Sponsor Awareness

In Hypothesis H.1.4 we propose that people who watch the sponsored event
more often will better recall the sponsors.

To test Hypothesis H.1.4 we proceed as above. Using a multinominal logistic
regression, we determine the effects of amount of exposure on sponsor recall (no
recall/aided recall/unaided recall). Our results confirm that exposure significantly
(p < 0,01) and positively affects sponsor recall for all six sponsors.

Interpreting parameter estimates, we find that exposure is highly significant (BMW,
Milka, Carlsberg, Telekom Austria, Tag Heuer, p < 0,01) or at least weakly signifi-
cant (Memphis, p < 0,1) in distinguishing between no recall and unaided recall.
Chances of remembering a sponsor unaided decrease by between 43 percent (for
Memphis) to 60 percent (for BMW) if exposure is reduced by one level. Exposure
also is a significant predictor of membership at group no recall compared to group
aided recall (p < 0,01 for Carlsberg, Memphis and Tag Heuer, p < 0,05 for BMW,
Milka and Telekom Austria). The odds of aided sponsor recall decrease by
between 30 percent (for Telekom Austria) to 58 percent (for Carlsberg) if exposure
is reduced. Therefore, we conclude Hypothesis H.1.4 is supported empirically.


Regression Model

As explained in Section 2.2, we expect first that the image of the sponsor after the
event (SIAFTER) is influenced by its image before the event (SIBEFORE), since attitudes
towards brands tend to be stable over time (c.f., H.2.1). The magnitude of this
effect is represented by parameter β 1 in model (1). Second, in line with the mere
exposure effect we propose that unaided and aided sponsor recall (SRUNAIDED,
SRAIDED) also affect post-event sponsor image positively (c.f., H.2.2; parameters β 2
and β 3 in model (1)). We also assume that if an image transfer from the Alpine Ski
World Championships to the sponsoring brand took place (c.f., H.2.3), we will
find a significant positive impact of event image (EI). This effect is further sepa-
rated into a baseline component (β 4 in model (1)) common to all sponsors, and
also into components (γ 1,…,γ 5 in model (1)) that reflect individual differences
which might be due to sponsorship leverage and/or other circumstances such as
industry type. Because of modeling issues (i.e. multicollinearity) we introduce
dummy variables (SD1,…, SD5) for all but one sponsor. Thus, strictly speaking, β 4
represents the magnitude of image transfer of Carlsberg, γ i the incremental gain
characterizing respective sponsors. The right half of Figure 1 displays a compre-
hensive representation of our model.

In more formal terms, we pool the data over all sponsors and assume a common
intercept β 0 and fixed effects of pre-event sponsor image (β 1) and sponsor aware-
ness (β 2, β 3), but shifting slope coefficients of event image. Hence, we end up
with the following model:

130                                                                      sbr 56 (2/2004)
                                                                                  Sport Sponsorships

Figure 1: Graphical representation of the sponsorship model


                                     3UH(YHQW 6SRQVRU ,PDJH         E 
                                                              E E       3RVW(YHQW 6SRQVRU ,PDJH
                                          6SRQVRU Awareness

    (YHQW ,QYROYHPHQW                                          E 
                                           (YHQW ,PDJH

                                                               0RGHUDWLQJ 6SRQVRUVSHFLILF (IIHFWV

 'ULYHUV RI 6SRQVRU Awareness                    ,PDJH 7UDQVIHU ± 5HJUHVVLRQ 0RGHO

                                                                             5          
SI AFTER = β 0 + β1 ⋅ SI BEFORE + β 2 ⋅ SRUNAIDED + β 3 ⋅ SR AIDED +  β 4 + ∑ γ i ⋅ SDi  ⋅ EI + ε (1)
                                                                           i =1         

Because of different numbers of observations per sponsor (c.f., Section 3.2) and
pragmatic reasoning, we postulate identical independently distributed error terms
ε and use OLS estimation. If we were to use a more rigorous procedure, such a
procedure might specify a complex error structure and use GLS calibration.

Finally, we propose a linear model, because we do not expect any dominating
second-order effects. This assumption is confirmed when we apply other non-lin-
ear curve estimation methods that do not improve estimation results.


The model fits the data well (F = 62,93, p < 0,01, df: 604). The independent vari-
ables explain a reasonable amount of variance in post-event sponsor image (R2 =
0,48). The most influential predictor of post-event sponsor image is, as expected,
pre-event sponsor image (c.f., Table 2). Sponsor awareness also has a significant
positive impact, and the effect is, as expected, stronger for unaided recall than for
aided recall. Event image also has a significant positive effect. However, we do
identify sponsor-specific influences. There seems to be one group of sponsors
(i.e., Carlsberg, Memphis) that benefits least from their sponsoring efforts. On the
other hand, BMW gains most from this promotional campaign with a combined
effect of 0,38 of event image on its own image. The other brands form the inter-
mediate group. Overall, we conclude that our results support Hypotheses H.2.1,
H.2.2 and H.2.3.

sbr 56 (2/2004)                                                                                         131
R. Grohs/U. Wagner/S. Vsetecka

Table 2: Estimated standardized regression coefficients
         (dependent variable: SIAFTER)

    Variables                                        ˆ      ˆ
                                                    β 1′,…,β 4′ a                   Significance
    SIBEFORE                                             0,49                          < 0,01
    SRUNAIDED c                                          0,19                          < 0,01
    SRAIDED c                                            0,08                          < 0,05
    EI                                                   0,18                          < 0,01
                                                    γ 1′,…,γ 5′ b
                                                    ˆ       ˆ
    SDBMW * EI d                                         0,20                           <   0,01
    SDMilka * EI d                                       0,10                           <   0,05
    SDTelekom Austria * EI d                             0,09                           <   0,05
    SDTag Heuer * EI d                                   0,12                           <   0,01
    SDMemphis * E d                                      0,08                           <   0,10
a     ˆ      ˆ
     β1′,…,β4′ are the estimates of the standardized regression coefficients of model (1).
b    γ 1′,…,γ 5′ are the estimates of the standardized regression coefficients of model (1).
      ˆ     ˆ
c    Coefficients for unaided and aided recall must be interpreted from the base level of “no recall”. Sig-
     nificant positive β2, and β3, respectively, thus indicate there is a positive effect of unaided and aided
     recall on post-event sponsor image compared with zero effects in the case of no recall.
d    Coefficients for sponsor specific dummy variables must be interpreted with respect to the base level
     sponsor Carlsberg. A significant positive γi thus indicates that for the corresponding sponsor the
     impact of event image is significantly stronger than for Carlsberg.



Sponsor Awareness

Sponsor recall depends on several factors. Empirically, we identify positive effects
of event-sponsor fit, event involvement, and exposure on sponsor recall. We do
not find consistent support for the hypothesis that brand prominence increases
sponsor awareness. Therefore, companies acting as sponsors can reduce the dan-
ger of ambush marketing by deliberately incorporating these factors into sponsor-
ship planning and execution.

Selecting those sponsorships for which event-sponsor fit, event involvement, and
exposure are high is likely to result in increased sponsor recall. Brand promi-
nence, the one parameter over which the sponsor has no control, only partly
affected sponsor recall. These observations support the conclusion that the choice
of an appropriate sponsorship makes a significant difference in sponsor recall, and
can be achieved by including the factors outlined in our study in the sponsorship
decision making process.

We also show that effects of event-sponsor fit, event involvement, and exposure
on sponsor awareness can be observed independently from sponsor’s industry
type. However, further research is necessary to determine whether our sports

132                                                                                          sbr 56 (2/2004)
                                                                              Sport Sponsorships

sponsorship-specific findings can be extended to the sponsorship of arts, social
activities, or the like.

Image Transfer

Our empirical study confirms that a general transfer of associations from the spon-
sored event to the sponsor takes place for all sponsors, regardless of industry
type. However, our results also show that sponsor specificities do matter. These
differences between sponsors demand further attention and investigation, because
they might be attributable to a sponsor’s additional promotional activities or to the
fit between event and sponsor.

Researchers agree that the sponsorship itself, i.e., paying the sponsorship fee, is
only a starting point. Parker 67 even states “… sponsorship without support is like
buying advertising space then having nothing to show. You end up with a blank
screen or an empty page.” Recent studies document empirically the importance of
communicating sponsorships. Crimmins/Horn 68 and Quester/Thompson 69 show
that the degree to which a sponsorship is leveraged has a significant positive
impact on sponsor image.

Therefore, we collect data on supporting activities undertaken by the six sponsors
of the Alpine Ski World Championships. We find the highest degree of leverage
for Milka and BMW, both of whom communicated their sponsorship by using
national advertisements on TV and on their website. We also find high leverage
for Telekom Austria, which used both national television and print media adver-
tisements. Furthermore, Milka and Telekom Austria carried out raffles on TV
before and after the ski races. But Carlsberg (with the exception of some informa-
tion about the World Championships on its homepage), Memphis, and Tag Heuer
did not leverage their sponsorship, at least not in Austria. Relating these observa-
tions to the magnitude of image transfer (i.e., γi′) shows that increased leverage
leads to a greater image transfer. This assumption seems to be supported particu-
larly in the case of Milka and BMW (large image transfer and high sponsorship
leverage) as well as for Carlsberg and Memphis (small image transfer and hardly
any activities supporting the sponsorship). Thus, our results indicate a positive
relation between sponsorship leverage and the magnitude of image transfer. How-
ever, interpretation of the data might be biased because of subjective judgments.
Further research needs to clarify and explore these issues with greater rigor.

Literature on classical conditioning states that conditioning is easier if there is a
certain fit between the unconditioned and the conditioned stimulus 70. Because
conditioning is also an important principle for sponsorships, with the sports event
being the unconditioned stimulus and the sponsor the conditioned stimulus, we
can assume that a better fit between sponsor and event leads to an increased
image transfer. This hypothesis is supported by empirical results from other

67   See   Parker (1990), cited according to Thwaites et al. (1998), p. 35.
68   See   Crimmins/Horn (1996).
69   See   Quester/Thompson (2001).
70   See   Rosenstiel/Neumann (1991).

sbr 56 (2/2004)                                                                              133
R. Grohs/U. Wagner/S. Vsetecka

authors 71, all of whom find that sponsors with a strong link to an event gain more
image out of the sponsorship than do other sponsors.

In our case of the Alpine Ski World Championships, there also seems to be a rela-
tion between the level of event-sponsor fit and the magnitude of image transfer.
Respondents in our study ranked the sponsors according to their perceived fit
with the event. Perceived fit is highest for Tag Heuer, followed by Milka, Telekom
Austria, BMW, Memphis, and Carlsberg. Image transfer was particularly high for
BMW, Milka, Tag Heuer, and Telekom Austria, and low for Memphis and Carls-
berg. Even though there seems to be some evidence to support our assumption,
the validity of our results is limited. A more stringent experimental design is
needed to elicit the true effects and to support or oppose our preliminary results.


Implications from our research include both general guidelines and specific sug-
gestions for managers who plan and execute sponsorships. Overall, it seems to be
important and necessary for a viable sponsorship to define targets well in terms of
brand awareness and/or image transfer. These goals are related to consumers’
behavioral intention more directly than are the exposure goals that are often used
in practice.

We have also demonstrated that pursuing these objectives still allows to attribute
the effects on sponsor awareness and image directly to the sponsorship, which is
not all that easy in the case of behavioral success measures such as sales. How-
ever, setting targets is not enough. Companies must also evaluate achievement by
using appropriate techniques. Although assessing sponsor awareness is straightfor-
ward, our paper offers a new method of assessing image transfer. Companies
might use the general framework outlined here to put their sponsorship evaluation
on a more systematic, solid basis.

Another important task for managers is to make sure that they consider the factors
affecting sponsorship success and that these factors are incorporated into the plan-
ning process. We present evidence that sponsorships are worthwhile for increas-
ing sponsor awareness and improving brand image, and are especially successful
if sponsors and sports marketers jointly design the favorable circumstances dis-
cussed in this paper.

In line with our results we propose to executives in sponsoring firms that they:

1. conclude contracts only if a certain amount of exposure is guaranteed;

2. select sports where there is high fan involvement and commitment to the spon-
   sored activity;

3. forge a link between the event and the sponsor by communicating the spon-
   sorship in order to increase perceived event-sponsor fit;

71 See Cornwell/Pruitt/van Ness (2001); d’Astous/Bitz (1995); Gwinner/Eaton (1999).

134                                                                                   sbr 56 (2/2004)
                                                                 Sport Sponsorships

4. make sure that leveraging the sponsorship is possible close to the event, for
   instance, by reserving advertising time for sponsors before and after broadcasts
   of the sponsored event.
Corresponding guidelines apply for sports marketers. In the process of sponsor
acquisition, sports marketers should:
1. try to ensure that exposure is high (for instance by signing coverage contracts
   with television channels) and use this information when negotiating sponsor-
2. approach sponsors that have the closest fit with the event;
3. in the long run, try to promote sports with a high fan involvement.
Overall, our research helps to clarify a number of important issues in sports spon-
sorship. We formalize and test the effects of a number of success factors on
achieving awareness and image goals for sponsoring companies, thereby making
the process of sponsorship decision making, planning, and execution more ratio-
nal and systematic. Particularly in the field of image transfer we find that event
image is an important driver of post-event sponsor image. Averaged across all
sponsors, the effect of event image on post-event sponsor image is approximately
half the size of pre-event sponsor image.
However, despite this convincing evidence, there are several issues that still need
to be examined in more detail. After finding support for an image transfer taking
place, the question arises: What are the drivers of the magnitude of image transfer?
Here, it would be worthwhile to develop a conceptual model that incorporates all
the factors that influence image transfer.
We presented first insights into the importance of event-sponsor fit as well as
sponsorship leverage, but we have not tested any hypothesis allowing its falsifica-
tion. In this area, we recommend a quantitative analysis of these two factors for
further research projects. We also believe that for event-sponsor fit, it would be
promising to test for additional industries for specific effects.
Another aspect we raised is the distinction between two ways of looking at the
issue of event-sponsor fit. On the one hand, we demonstrated that there are
industry-specific effects. The more similar a brand is in relation to an event, the
larger the image transfer. In this case, we categorize fit and measure it across
brands. On the other hand, we assessed the effect of person-specific fit for each
sponsor separately. A person who gave a higher rating to his or her perceived fit
between event and a specific sponsor recalled the sponsor better than did those
respondents who perceived a lower fit. This finding is consistent for all sponsors,
with the fit being assessed within brands but across respondents.
Finally, there is one more task which needs further consideration and is likely to
stimulate research: We tested our hypotheses at only one sports event. For spon-
sors it would be instructive to know if effects we found generalize to other sport
events and whether different associations, such as those conveyed by art sponsor-
ships, can be passed on to the sponsor in a similar fashion.

sbr 56 (2/2004)                                                                 135
R. Grohs/U. Wagner/S. Vsetecka


Questionnaire items, translated                                  Response categories
Sponsor Recall (unaided/aided)
– Which sponsors of the Alpine Ski World Championships 2001      Open question
  in St. Anton do you recall?
– Which companies in the following industries were in your       Open question
  opinion Sponsors of the Alpine Ski World Championships
  automobile, chocolate, telecommunication, watch, cigarettes,
Brand Prominence
– Are you familiar with the company/brand “sponsor name”?        1 … yes
                                                                 2 … no
  BMW, Milka, Telekom Austria, Tag Heuer, Memphis, Carlsberg
Event-Sponsor Fit
– I see a link between the Alpine Ski World                      1   …   certainly
  Championships and “sponsor name”.                              2   …   rather yes
                                                                 3   …   partly
– My associations with the Alpine Ski World                      4   …   rather not
  Championships are similar to those I have with                 5   …   not at all
  “sponsor name”.
– The ideas that come to my mind when asked about
  “sponsor name” are related to the ideas I have about
  the Alpine Ski World Championships.
Event Involvement
– How would you describe your interest in the Alpine Ski         1   …   very high
  World Championships 2001 in St. Anton?                         2   …   high
                                                                 3   …   average
                                                                 4   …   low
                                                                 5   …   no interest

– How often did you watch the races of the Alpine Ski World      1   …   daily
  Championships 2001 on TV?                                      2   …   > 5 times
                                                                 3   …   3-5 times
                                                                 4   …   < 3 times
                                                                 5   …   not at all

Pre-Event and Post-Event Sponsor Image, Event Image
– How do you perceive the relation between the Alpine Ski        1   …   very strong
  World Championships in St. Anton and the following             2   …   strong
  concepts?                                                      3   …   average
                                                                 4   …   weak
– sport, dynamic, health, success, Austria, tradition, youth,    5   …   no relation
  precision, risk, sociableness, enthusiasm, winter

136                                                                          sbr 56 (2/2004)
                                                                               Sport Sponsorships


Abratt, Russell/Clayton, Brian C./Pitt, Leyland F. (1987), Corporate Objectives in Sports Sponsorship, in:
  International Journal of Advertising, Vol. 6, pp. 299 – 311.
Backhaus, Klaus/Erichson, Bernd/Plinke, Wulff/Weiber, Rolf (2003), Multivariate Analysemethoden –
  Eine anwendungsorientierte Einführung, 10th ed.
Churchill, Gilbert A./Iacobucci, Dawn (2002), Marketing Research: Methodological Foundations, 8th ed.
Cornwell, Bettina T./Maignan, Isabelle (1998), An International Review of Sponsorship Research, in:
  Journal of Advertising, Vol. 27, pp. 1 – 21.
Cornwell, Bettina T./Pruitt, Stephen W./van Ness Robert (2001), The Value of Winning in Motorsports:
  Sponsorship-linked Marketing, in: Journal of Advertising Research, Vol. 41, pp. 17 – 31.
Cornwell, Bettina T./Roy, Donald P./Steinard, Edward A. II (2001), Exploring Managers’ Perceptions of
  the Impact of Sponsorship on Brand Equity, in: Journal of Advertising, Vol. 30, pp. 41 – 51.
Crimmins, James/Horn, Martin (1996), Sponsorship: From Management Ego Trip to Marketing Success,
  in: Journal of Advertising Research, Vol. 36, pp. 11 – 21.
d’Astous, Allen/Bitz, Pierre (1995), Consumer Evaluations of Sponsorship Programmes, in: European
  Journal of Marketing, Vol. 29, pp. 6 – 22.
Deimel, Klaus (1992), Wirkungen der Sportwerbung.
Drees, Norbert (1992), Sportsponsoring, 3rd ed.
Dreyer, Axel (1986), Werbung im und mit Sport.
Glogger, Anton (1999), Imagetransfer im Sponsoring: Entwicklung eines Erklärungsmodells.
Gwinner, Kevin P. (1997), A Model of Image Creation and Image Transfer in Event Sponsorship, in:
  International Marketing Review, Vol. 14, pp. 145 – 158.
Gwinner, Kevin P./Eaton, John (1999), Building Brand Image Through Event Sponsorship: The Role of
  Image Transfer, in: Journal of Advertising, Vol. 28, pp. 47 – 57.
Hansen, Flemming/Scotwin, Lene (1995), An Experimental Enquiry into Sponsoring: What Effects can
  be Measured?, in: Marketing and Research Today, pp. 173 – 181.
Harvey, Bill (2001), Measuring the Effects of Sponsorships, in: Journal of Advertising Research, Vol. 41,
  pp. 59 – 65.
Hermanns, Arnold (2000), Sponsoring Trends 2000.
Javalgi, Rajshekhar G./Traylor, Mark B./Gross, Andrew C./Lampman, Edward (1994), Awareness of
  Sponsorship and Corporate Image: An Empirical Investigation, in: Journal of Advertising, Vol. 23, pp.
  47 – 58.
Johar, Gita V./Pham, Michel T. (1999), Relatedness, Prominence, and Constructive Sponsor Identifica-
  tion, in: Journal of Marketing Research, Vol. 36, pp. 299 – 312.
Keller, Kevin L. (1993), Conceptualizing, Measuring, and Managing Customer-Based Brand Equity, in:
  Journal of Marketing, Vol. 57, pp. 1 – 22.
Kroeber-Riel, Werner/Weinberg, Peter (2003), Konsumentenverhalten, 8th ed.
Lardinoit, Thierry/Derbaix, Christian (2001), Sponsorship and Recall of Sponsors, in: Psychology &
  Marketing, Vol. 18, pp. 167 – 190.
McDonald, Colin (1991), Sponsorship and the Image of the Sponsor, in: European Journal of Market-
  ing, Vol. 25, pp. 31 – 38.
Meenaghan, Tony (1991), The Role of Sponsorship in the Marketing Communications Mix, in: Interna-
  tional Journal of Advertising, Vol. 10, pp. 35 – 47.
Meenaghan, Tony (2001), Understanding Sponsorship Effects, in: Psychology & Marketing, Vol. 18, pp.
  95 – 122.
Meerabeau, Elizabeth/Gillett, Roy/Kennedy, Michael/Adeoba, Johnson/Byass, Michael/Tabi, Kingsley
  (1991), Sponsorship and the Drinks Industry in the 1990s, in: European Journal of Marketing, Vol.
  25, pp. 39 – 56.
Otker, Ton/Hayes, Peter (1987), Judging the Efficiency of Sponsorship: Experience from the 1986 Soccer
  World Cup, in: European Research, Vol. 15, pp. 3 – 8.
Pham, Michel T./Johar, Gita V. (2001), Market Prominence Biases in Sponsor Identification: Processes
  and Consequentiality, in: Psychology & Marketing, Vol. 18, pp. 123 – 143.

sbr 56 (2/2004)                                                                                      137
R. Grohs/U. Wagner/S. Vsetecka

Quester, Pascale/Farrelly, Francis (1998), Brand Association and Memory Decay Effects of Sponsorship:
   The Case of the Australian Formula One Grand Prix, in: Journal of Product & Brand Management,
   Vol. 7, pp. 539 – 556.
Quester, Pascale/Thompson, Beverley (2001), Advertising and Promotion Leverage on Arts Sponsorship
   Effectiveness, in: Journal of Advertising Research, Vol. 41, pp. 33 – 47.
Rajaretnam, John (1994), The Long-Term Effects of Sponsorship on Corporate and Product Image:
   Findings of a Unique Experiment, in: Marketing and Research Today, pp. 62 – 74.
Rosenstiel, Lutz/Neumann, Peter (1991), Einführung in die Markt- und Werbepsychologie, 2nd ed.
Sandler, Dennis M./Shani, David (1989), Olympic Sponsorship vs. “Ambush” Marketing: Who Gets the
   Gold, in: Journal of Advertising Research, Vol. 29, pp. 9 – 14.
Scott, D.R./Suchard, Hazel T. (1992), Motivations for Australian Expenditure on Sponsorship – An
   Analysis, in: International Journal of Advertising, Vol. 11, pp. 325 – 332.
Stipp, Horst/Schiavone, Nicholas P. (1996), Modeling the Impact of Olympic Sponsorship on Corporate
   Image, in: Journal of Advertising Research, Vol. 36, pp. 22 – 27.
The Wall Street Journal (1998), Public Is Confused on Olympic Sponsors, February 18th, B8.
Thjomoe, Hans M./Olson, Erik L./Bronn, Peggy S. (2002), Decision Making Processes Surrounding Spon-
   sorship Activities, in: Journal of Advertising Research, Vol. 42, pp. 6 – 15.
Thwaites, Des (1995), Professional Football Sponsorship – Profitable or Profligate, in: International Jour-
   nal of Advertising, Vol. 14, pp. 149 – 164.
Thwaites, Des/Aguilar-Manjarrez, Rafael/Kidd, Callum (1998), Sports Sponsorship Development in
   Leading Canadian Companies: Issues and Trends, in: International Journal of Advertising, Vol. 17,
   pp. 29 – 49.
Zajonc, Robert B. (1968), Attitudinal Effects of Mere Exposure, in: Journal of Personal and Social Psy-
   chology, Vol. 9 (Pt. 2).

138                                                                                       sbr 56 (2/2004)

Shared By: