An analysis of the influence of discount sales promotion in consumer buying intent and the moderating effects of attractiveness

O objetivo neste trabalho foi analisar a influência da promoção de vendas de desconto na intenção de compra e os efeitos moderadores da atratividade nas relações entre intenção de compra de um produto em desconto e a impulsividade, percepção hedônica e de risco financeiro. Para isso, um experimento com a participação de 613 estudantes foi realizado. As hipóteses previam que um produto com promoção de desconto se relacionaria positivamente com a impulsividade e com a percepção hedônica sobre o bem ofertado, e ainda, negativamente, com a percepção de risco financeiro associado ao produto em desconto. Esperava-se uma moderação positiva da percepção da atratividade da promoção de desconto divulgada sobre as intenções dos comportamentos. Os resultados confirmaram as hipóteses, indicando efeitos positivos da impulsividade e percepção hedônica com a compra do produto em desconto, além do vínculo negativo existente entre a intenção de compra do produto em desconto e a percepção de risco financeiro. Os efeitos moderadores não foram confirmados. As considerações finais encerram o trabalho.


INTRODUCTION
Sales promotion techniques are instruments that seek to increase sales of products and brands, usually in a short time (Wierenga & Soethoudt, 2010), because they act in the consumer's mind as a benefit to him, creating thus consumer behavior (Yusuf, 2010).The effectiveness and the importance of sales promotion in the market can be viewed when presenting the segment numbers.According to Teunter (2002), over 20% of sales of products of some food branches occur through sales promotion activities.In a report quoted by Wierenga and Soethoudt (2010), over 75% of spending on communication in nondurable consumer goods segment between 1997 and 2004 in the United States were driven by sales promotion activities, while 25% were applied in other communication activities.
In addition to market fundamentals, the importance is attested in the scientific field, whose studies have been conducted for several decades.In 1978, Preston, Dwyer, and Rudelius already demonstrated the effect of sales promotion to convince consumers to open a bank account.In that study, it was found that 50% of the increase in number of accounts in a bank searched occurred during a promotional campaign.More recently, D' Austous and Jacob (2002) demonstrate the efficiency of this promotional tool to influence the computer purchase, while Boschetti (2012) confirmed the positive linkages, promoting the behavior of purchase of financial services.
To achieve the proposed objectives, this work was based on a laboratory experiment.

THEORETICAL BASIS AND ASSUMPTIONS OF THE STUDY
According to the studies of Campbell and Diamond (1990), D'Astous and Landreville (2003) and Alvarez and Casielles (2005), sales promotions are distributed into two groups: price or monetary promotions (i.e.discounts, coupons, rebates); non--monetary promotions or not geared to price (i.e.free samples, distribution of promotional prizes, gifts, contests and loyalty programs).
Sales promotion techniques are instruments that seek to increase sales of products and brands, usually in a short time (Wierenga & Soethoudt, 2010), because they act in the consumer's mind as a benefit to it, thus generating a consumer behavior (Yusuf, 2010).Therefore, it is believed that sales promotion has a positive influence on the purchase intention.Nevertheless, studies such as those conducted by Cotton and Babb (1978), Lattin and Bucklin (1989) and, more recently, Alvarez and Casielles (2005) have demonstrated the effectiveness of sales promotions in the responses and buyers' attitudes, both monetary as non-monetary, over the past decades.For this reason, it is believed that: H 1 : The discount promotion positively influences consumer purchase intent.
Another assumption is the relationship between impulsivity and intention to buy a product on discount.Impulsivity is associated with emotional and hedonic components of personality (Rook 1987;Rook & Fisher, 1995;Weinberg & Gottwald, 1982;Youn, 2000).In the present research, it is assumed that consumer impulsiveness will influence the intention of buying discounted products, since we detected a relationship between short-term satisfaction of impulsive consumers (Dholakia, 2000;Youn & Faber, 2000) and the benefits provided to consumers who purchase products under the effects of monetary promotions -case of discount campaigns, which have immediate impact, i.e. also a short-term effect (Alvarez & Casielles, 2005;Blattberg & Neslin, 1990).Therefore, the following hypothesis arises: H 2 : Consumer impulsiveness positively influences the intention of buying a product on discount.
The next hypothesis suggests a relationship between hedonic perception of the offered product and purchase intent with discount.Hirschman and Holbrook (1982) introduced the concept of hedonic consumption and related it to the multi--sensorial and emotional involvement, which, in turn, is linked to the experiences of consumers and products.The size of hedonic consumption can be derived from the exclusivity of the good, its symbolic meaning, the emotional excitement and imaginary evocation (Hirschman & Holbrook, 1982;Spangenberg, Voss, & Crowley, 1997).The hedonic dimension is more subjective and personal than the utilitarian one (Babin, Darden, & Griffin, 1994).The motivation of hedonic consumption derives from a need for pleasure and excitement that can be provided by these products (Dhar & Wertenbroch, 2000).The pleasure pathways Fernando de Oliveira Santini, Cláudio Hoffmann Sampaio, Marcelo Gattermann Perin and Valter Afonso Vieira are the real motivations of hedonic consumption (Ramanathan & Menon, 2006).Inman, McAlister and Hoyer (1990), Schindler (1992) and Dhar and Hoch (1996) show that it is not only financial gain that is related to sales promotion.Therefore, the emotional effects and, in this case, hedonic benefit, may also help to explain, for example, why consumers respond more strongly to coupon campaigns in stores than to advertisements announcing a price reduction (Schindler, 1992;Dhar & Hoch, 1996) and also because consumers respond to negligible reduction in price (Hoch, Drèze, & Purk, 1994;Inman et al., 1990).High value purchases may also be related to the dimension of choice optimization proposed by Westbrook and Black (1985), since the fact of finding a discount or a bargain may lead to satisfaction of personal achievement, a characteristic of hedonistic consumption (Hirschman & Holbrook 1982;Spangenberg et al., 1997).Based on this evidence, we propose the next hypothesis: The hedonic perception of a product positively influences the purchase intention of discounted good.
The perceived risk has been shown to be an important construct in consumer behavior explanation and its research has been widely discussed (Blackwell, Miniard, & Engel, 2005;Solomon, 2011).Not only in the marketing field, but also in the field of psychology, the debate on this issue has been highlighted (Dholakia, 2000;Mitchell, 1999).Dholakia (2000) sets out the general concept of risk perceived as a negative sentiment that may be generated through the acquisition of a product/service.Researches such as those carried out by Hisrich, Dornoff and Kernan (1972) and Choi and Lee (2003) show that the feeling of perceived risk affects consumer propensity to make a decision when buying.In the context of risk theory, people assess the risk before making a decision (Weegels & Kanis, 2000) and the feeling of risk is increased in situations that are beyond the daily experience.
One of the dimensions of perceived risk is associated with the financial risk and occurs when consumer's sentiment is linked to monetary loss that the good to be acquired may cause (Dholakia, 2000;Huang, Schrank, & Dubinsky, 2004;Jacoby & Kaplan, 1972).This loss may be linked to a simple devaluation of the product or to expenditures on repairs and maintenance (Apaydin & Köksal, 2011;Mitchell, 1999), and is one of the most significant dimensions in determining that feeling (Agarwal & Teas, 2001;Mitchell, 1999;Ofir & Bechtel, 1990).Similarly, research efforts have been conducted to specifically analyze this dimension of perceived risk (Diacon, 2004;Engelberg, 2007;Simon & Victor, 1994).In this context, a relationship that may be associated with sales promotion and a greater strength to action campaigns with monetary goal arises, such as, for example, discount, since this tool intends to reduce the traditional value of the offered product (Alvarez & Casielles, 2005;Bawa & Shoemaker, 1987;Blattberg & Neslin, 1990;Gupta, 1988;Lee, 2002;Winer, 1985).This relationship is explained by Tan and Chua (2004) when they say that monetary promotions are evaluated by consumers as a loss reduction.Similarly, Simon and Victor (1994) state that monetary incentives are efficient alternatives for consumers to decrease the perception of risk.This link gets partial confirmation in the study of Garretson, Burton, and Clow (1999) because, on the one hand, the promotion can encourage consumption, and on the other can also depreciate the perception and quality of a product and thus increase feelings associated with risk.In this logic, the discount promotion could promote a sense of prudence (Chandon, Wansink, & Laurent, 2000;Kwok & Uncles, 2005) and, consequently, relate negatively to the purchase intent of good discount.Therefore, it is proposed that: The perception of financial risk on a product influences negatively the purchase intention of a discounted good.
It is expected that the perceived attractiveness of the promotion is a moderating effect of the last three suggested hypotheses.The attractiveness of the promotion is emphasized by Simonson et al. (1994), andD'Austous andLandreville (2003) and Liao (2006) as a guiding factor in the success or failure of a respective promotion.Researches made by Santini (2008) and Santini et al. (2011) showed that the attractive prize positively interferes in purchase intent, with a direct relationship.In the 1990s, Simonson et al. (1994) concluded that promotional campaigns with attractive prizes cannot be harmful to the brand image and attitude towards it.Therefore, it is not enough to make a promotion; the factor that can lead to the success or failure of the action is the level of its attractiveness.Indeed, it is assumed that an attractive prize encourages the purchase of products/services when consumers are not sure about their preferences in a class of products, amplifying the previously existing effect of discount promotion, impulsiveness, hedonism and reducing risk.
After nearly ten years of Simonson's et al. research (1994), D'Astous andJacob (2002) sought to understand consumer reactions across promotional offers.After conducting a qualitative and quantitative research, they found that consumers were interested in awards that were delivered upon purchase, presenting the value of the benefit and representing an attractive gift.A year later, D'Astous and Landreville (2003) ratified parts of the results, having identified positive relationships between the attractiveness of a promotion and evaluation of the brand of the product in promotion.In this study, an analysis of the relationship of the award with the product on offer was made.It turned out that, when the award is attractive, regardless of whether it has a connection or not with the product offered, the campaign is well evaluated.On the other hand, an unattractive award is badly evaluated when it has no connection with the product on offer.It was also possible to verify that, if the purpose of a promotion is to arise consumer interest, the award has to be attractive.Therefore, it is believed that the award (attractive or unattractive) is a significant intervening variable to explain the reactions of consumers about a promotional campaign, acting as moderator.
Other empirical evidence arise to the attractiveness moderation hypothesis in the three suggested direct relations.First, Liao (2006) noted that the benefit distributed in a promotion plays a key role in consumer perception of the product brand, and the attractiveness of the promotion is an essential factor for positive assessment of the brand offered.Second, Palazón and Delgado-Ballester (2009) investigated the interaction of promotion attractiveness between monetary and non-monetary campaigns.The results demonstrate that the attractive discount promotion influences positively on purchase intent, while the low attractiveness is less effective in this behavior.Thus, it is expected that the discount attractiveness interfere positively in the suggested relationships, including the relationship between perceived risk and purchase intent.Therefore: H 5 : The perception of the attractiveness of the sales promotion positively moderates the relationship between (H 5a ) consumer impulsivity and purchase intent, (H 5b ) hedonic perception and purchase intent, and (H 5c ) financial risk and purchase intent.
Figure 1 shows the four antecedents of purchase intent and the moderating role of attractiveness promotion interacting with the hypothesized direct relations.

METHODOLOGICAL PROCEDURES
At first exploratory and descriptive surveys were conducted, whose purpose was to prepare for the experiment conducted in this study.The exploratory stage guided the identification of possible products and discounts in line with the reality of the participant public of the study.These products and discounts were later validated by the views of the participant sample in the descriptive stage and then were handled in the experimental stage.
As for the exploratory stage, the collection was based on secondary data -the collective purchasing website Groupon, from where we analyzed the products and discounts offered during the period of twenty days of July 2012.From the execution of this analysis it was possible to evaluate the ten more frequent products under promotion and the average of the discount offered in that collective purchasing site.
Once we identified the products and discounts obtained in the exploratory stage, we began descriptive stage, culminating in the consolidation of the more familiar and realistic product in sight of the participants, since this initiative is essential to internal and external validity of an experiment laboratory (Wilson, Aronson, & Carlsmith, 2010).

Fernando de Oliveira Santini, Cláudio Hoffmann Sampaio, Marcelo Gattermann Perin and Valter Afonso Vieira
Before using the research instrument of the descriptive stage, we made the validation of the questionnaire content and the pretesting.Three academics carried out the content validation, suggesting the exclusion of three products, because they were not aligned with the reality of respondents -lamp, high-pressure washer and wedding rings -plus the videogame product, since it is a good referring to a specific brand, for instance X-Box and PlayStation.The pre-test together with fifteen (15) students who completed questionnaires on different types of products, not demanded adjustments to the scales used, nor the revaluation of advertised products.
For the descriptive stage, the constructs of the survey instrument were operationalized from the scales of: interest on the product, from Bruner and Hensel (1998), whose aim was to analyze the buying interest in relation to the publicized product; hedonic or utilitarian sense, from Voss, Spangenberg, and Grohmann (2003); financial risk, from Stone and Gronhaug (1993).The questionnaires were presented to 182 students from the Business Administration course.For each tested product (camera, tablet, bicycle, digital book reader, netbook, mobile device), the stipulated sample was of at least thirty valid questionnaires.It should be noted that this research aims to evaluate a single product from the perspective of the feelings associated with it by surveyed consumers and, therefore, the intention is not to make the manipulation of concepts from the product perspective but from the consumer perspective.In this sense, we considered not to choose products that arouse an extreme behavior, since this fact could interfere in the search results.Moreover, a product that brings high perceived financial risk would tend to inhibit a purchase intent, regardless of a discount or not (Sarin, Sego & Chanvarasuth, 2003;Teimoury, Fesharaki, & Bazyar, 2010).
In this sense, the product used in the experimental stage was based on behavior observation with a less predominant tendency (purchase intent, hedonic perception, utilitarian perception and financial risk), obtained by averaging analysis, as well as thinking on the variation of participants responses to every behavior, which was analyzed from the standard deviation technique.The result presented in this step identified the netbook product (Figure 2) as the product with more changes in behavior (hedonic perception, utilitarian, purchase intent and financial risk).Once the product to be used was identified, the average discount the said website provides for this product was evaluated.This information was obtained in the exploratory stage, according to the details already exposed.Making this assessment, we came to the discount percentage of 37%, which was used in the experimental stage.

Population and survey sample
The study population is made of students from a university course in business administration, totaling approximately 1,500 students.The sample of the study was 662 students, selected for convenience method.

Experimental design
In the case of this research, we developed a laboratory experiment in which we handled the promotion factor (with discount promotion vs. no discount promotion).We chose to use

Figure 2: Product With a Promotion
the within-subjects method, in which each participant sample is subject to all experimental conditions (Keppel, 1991).
The manipulated factor was the promotion (discount vs. no discount).For this, we first delivered to surveyed people an ad of the selected product in the previous steps (exploratory and descriptive).This ad made a brief description of the product, with its approximate value (collected from the exploratory stage) and brought five questions to assess the respondents' intention of buying.
Once completed, the questionnaire was collected and the participant received once more the announcement of the same product with the same characteristics, but now with the highlight of the discount percentage promotion and the new product value.Besides this information, respondents were submitted to purchase intent scales (Bruner & Hensel, 1998), hedonic perception (Voss et al., 2003), financial risk (Stone & Gronbaug, 1993), impulsiveness (Puri, 1996) and attractiveness of the promotion (D'Astous & Landreville, 2003).

Control of external variables
In order to avoid external variables in the proposed experiment, the product offered did not contain any brand.This decision and criteria aims to avoid the influence of the brand in purchase intent (Keller & Lehmann, 2006), in the attitude (D'Astous & Landreville, 2003) and in the loyalty (Laroche, Pons, Zgolli, Cervellon, & Kim., 2003;Nagar, 2009) of the participant and increase the validity of the work.
It sought also to minimize the effects of sociodemographic influences in the search results, for example income, gender and age, since the research with students allows a certain homogeneity of the sample and an increase in the internal validity of the experiment (Peterson, 2001).

RESULTS
In order to prepare data for subsequent analysis, we made a purification in the database, excluding questionnaire respondents considered outliers.Moreover, in this analysis stage, we certified that the data meet the basic assumptions of multivariate data analysis (Kline, 1998).
Initially, frequency analysis was performed for all survey questions.This procedure aims to observe if there was a very large number of non-response to a question of the survey instrument.At this stage, no question was eliminated, since there was not a significant number of non-response (greater than ten), by variable.Then we proceeded to the evaluation of the questionnaires with more than 10% of unanswered questions.At this stage 44 (forty-four) questionnaires were eliminated.
Right after this step, we mad the analysis of univariate and multivariate outliers, according to the procedures outlined in the method.First, to assess the univariate outliers, the Z scores were calculated and five (5) cases with values greater than |3| for each variable were identified (Hair, Black, Babin & Anderson., 2005), which were excluded from the final analysis.In the case of multivariate outliers' analysis, we adopted the Mahalanobis calculation.In this situation, no case was identified, since we did not detect a large distance between the individual value and the obtained sample averages (Jöreskog, 1993;Kline, 1998).From there, we made the analysis of the characteristics of the sample.

Sample characterization
Based on the total universe of valid questionnaires, which were six hundred and thirteen elements, we made the descriptive statistical analysis in order to understand the characteristics of the sample survey.Briefly, a sample was obtained with a higher concentration of individual monthly income between R$ 1,000 and R$ 3,000 (53.9%), aged 19-25 years (48.9%),male (53.4%).

Test and discussion of hypothesis H 1
First, by t-test analysis we performed an overall comparison of the sample between the average purchase intent of the netbook product, excluding the discount promotion effect and, subsequently, the purchase intent with the effect of the promotion.According to Table 1, it is possible to observe that when the netbook product was submitted without the discount information, the average purchase intention was 2.92; this average rises to 3.45 (an increase of 18%; p <0.001) when the same ad product was presented with the discount promotion as sole change.Thus, the results confirm the hypothesis H 1 .
The first hypothesis of this study showed that the discount promotion positively influences consumer purchase intent.This  , 2005;Gupta, 1988), including Brazil (Boschetti, 2012;Santini et al., 2011), as well as strengthens the theoretical assumptions that the sales promotion is an important element to influence purchasing behavior of a customer (Blattberg & Neslin, 1990), stimulating consumer demand (Santini & Espartel, 2010) and encouraging brand change (Davis, Inman, & McAlister, 1992;Gupta, 1988).
In management terms, the finding confirms important elements in inventory management, by demonstrating that the discount promotion can be a key element to the stored product sales incentive and a differentiation tool in highly competitive product categories (Jones, 2008).In addition, managers can evaluate the use of this promotional tool to encourage the testing of new products (Nbudisi & Moi, 2005), raising the level of purchase of goods under promotion (Cotton & Babb, 1978) and increasing sales volume in a short-term (Blattberg & Neslin, 1990).

Discussion of measurement model
To test the hypothesis H 2 to H 5 , we used the structural equation modeling approach.In this regard, initially, we performed the analysis measurement model.At first, we evaluated the assumptions of normality of the observed variables.Thus, we made the kurtosis and skewness indicators calculation.Accordingly, we evaluated the kurtosis by means of Mardia coefficient, being considered significant when values lower than 5 were detected.Therefore, we detected the presence of normality in the distributions (Bentler, 1995;Mardia, 1970), since the displayed result was 1.93.In the asymmetry analysis, the observed variables showed a moderate symmetry, since Pearson' coefficient of skewness were close to zero.
Once the stage of normality analysis had ended, we did the confirmatory factor analysis, seeking to ratify the model originally proposed in theory.Thus, we developed the convergent and discriminant analysis, allowing us to check the reliability of each construct and the discrimination between them.
The factor loadings of the items in each variable were positive and significant (p <0.001).For the analysis of convergent and discriminant validity of the constructs, we used Fornell and Larcker's method (1982), where it was possible to identify that the average variance extracted (AVE) were always higher than the shared ones, which confirms such validation.Looking at the AVE results of each construct in the matrix table of the correlation between the constructs (Table 2), these are always higher than the correlations between the constructs, all of them being significant, which characterizes multicollinearity (Hair et al., 2005;Schumacker & Lomax, 1996).
We also assessed the validity of each composite construct.We observed that all the data was above the desirable reliability index (impulsiveness = 0.792; financial risk = 0.879; discount purchase intent = 0.954; perception hedonic = 0.952) and the factor loadings of the variables are greater than 0.50.

Test and discussion of hypotheses H 2 to H 4
Once past the model validation, we made the analysis of integrated models that aggregates the measurement model and the structural model, aiming to examine the influence of impulsiveness (H 2 ) perception Hedonic (H 3 ) and financial risk (H 4 ) on discounted product purchase intent.This review was carried out from the adjustment of the models and the statistical significance of the estimated coefficients (Kline, 1998).The adjustments of the models were as follows: Chi-square by Degrees of Freedom = 3.45; Comparative Fit Index = 0.965; Normed Fit Index = 0.951; and Root Mean Squared Error of approximation = 0.063.Table 3 presents the standardized coefficients and significance of the integrated model.
H 2 predicted a positive relationship between impulsivity and the purchase intention of a discounted product.The results confirmed the assumption (β = 0.202, p < 0.001).Indeed, this confirmation is consistent with the theoretical propositions suggesting the difficulty of impulsive control over the unwillingness to delay gratification that, in this case, are illustrated by the discount promotion (Hausman, 2000;Puri, 1996;Ramanathan & Menon, 2006;Rook, 1987).Confirmation of the result suggests some interesting managerial implications.
As mentioned, situational factors can pose significant results in impulsive purchase behavior (Mattson & Dubinsky, 1987).Thus, discount campaigns can bring an increase in sales volume resulting from purchases made mainly by people of impulsive character.It is suggested, too, based on the proposal from Rook and Fisher (1995), that a big part of consumers is impulsive, and the discount action may be a stimulator of sales increase, since people with impulsive characteristics have greater sensitivity to receive a reward, illustrated in this case by the discount promotion (Ramanathan & Menon, 2006).
Regarding H 3 , we managed to prove a positive relationship between the hedonic perception of the offered product and the purchase intention of the discounted good (β = 0.216; p <0.001).Thus, the analysis enhances the theoretical proposals that the discount promotion can be an inducer for the pursuit of personal gratification (Babin et al., 1994;Forsythe, Butler, & Kim, 1991), excitement (Hirschman & Holbrook, 1982;Spangenberg et al., 1997) and making unplanned purchases (Batra & Ahtola, 1990;Ramanathan & Menon, 2006).In the managerial context, the result may be useful to explain why, often, consumers respond to the insignificant reduction of prices for certain products (Hoch et al., 1994;Inman et al., 1990).This suggests to managers of products that have a more hedonic nature the possibility of working with small margins of discount.The result also suggests that the discounts used for products with more hedonic characteristics can cause increased excitement (Hirschman & Holbrook, 1982).Another interesting assumption is linked to the use of this tool for launching new products, given that this feeling is related to the hedonic perception (Kim, 2008;Venkatraman & Price, 1990) and the results showed positive relationships of this perception with the intent purchase of a discounted product.
As to H 4 , the proposal was that the perception of financial risk on a product would relate negatively with the intention of buying it with a discount (β = -0.221,p <0.001).This assumption was confirmed.Under the academic perspective, this result reinforces findings that this feeling is associated negatively with consumer purchase intent, a topic that has been the subject of repeated studies and assumptions for a long time (Blackwell et al., 2005;Solomon, 2011).Likewise, it enhanced the concept about fear of monetary loss linked to the feeling of perceived financial risk, even under the effects of a discount (Dholakia, 2000;Huang et al., 2004;Jacoby & Kaplan, 1972).For managers, this finding is interesting, in order for them to think about alternatives in addition to the discount promotion that may mitigate a negative perception of financial risk.Alternatives perhaps guided in product performance characteristics or in building and strengthening a brand, and not only in short-term actions, whose purpose is the sale, as is the case of discount campaigns (Alvarez & Casielles, 2005;Gupta, 1988).
Once the hypothesis of relationship between the constructs and purchase intent, as well as the adjustment of the integrated model are confirmed, the next stage of the work is the analysis of the last hypothesis of the research, suggesting the moderating effect of the discount attractiveness in the considered relations.Finally, it should be noted that, although it was no hypothesized as a direct effect, the attractiveness of the promotion is directly related to purchase intent (β = 0.59, t = 18.49, p <0.000).

Analysis and discussion of the moderating effect of the discount attractiveness (H 5 )
In the perspective of structural equation modeling analysis, it is possible to establish the categorization of a variable (nominal or ordinal) into two or more groups that are classified as moderator hypothesis (Krüll & McKinnon, 2001).In the case of this research, the moderator hypothesis proposes a difference in the intention of the paths between the constructs established because of the intensity related to the perception of the attractiveness of the discount promotion.In this case, the scale used to measure this behavior was transformed into dichotomous for two groups -high and low perception of

Fernando de Oliveira Santini, Cláudio Hoffmann Sampaio, Marcelo Gattermann Perin and Valter Afonso Vieira
attractiveness.Once separation of the groups was made, we estimated the parameters simultaneously to each group as standard coefficient rates (Table 4).Hypothesis H 5a predicted that the attractiveness would play a moderating role, specifically in a positive way in the relationship between impulsivity and purchase intent of a discounted product (H 5a ).It was expected that, in the group with higher attractiveness, the relationship between impulsivity and purchase intent of the discounted product would be higher compared to the lower attractiveness group.The analysis results found that the relationship between the two constructs was significant only for the low perceived attractiveness group.Moderation was considered significant because there have been major changes in the analyzed groups, but contrary to what was expected.
Increasing the understanding about the hypothesized relationship, we followed the Aiken and West (1999) procedure, which uses the nonstandard coefficient in a plot to display the interaction (Figure 3).In this case, impulsiveness values were calculated for high (+1 standard deviation) and low scores (-1 standard deviation) as a function of high (+1 standard deviation) and low values (-1 standard deviation) of moderating attractiveness (Aiken & West, 1999).Therefore, the slope and the intercept of the impulsiveness regression in purchase intent depends on the values assumed by the moderating attractiveness (West, Aiken, & Krull, 1996).

Figure 3: Moderating Attractiveness of the Relationship Between Impulsivity and Purchase Intent
The non-confirmation of the hypothesis may be related to the scale used in the study (Puri, 1996).This is because the measure proposed measures broadly consumer impulsiveness and not only an impulsive purchase.Accordingly, the scale has two dimensions -one related to issues of pleasure orientation, also highlighted by the author as hedonic characteristics of the individual; and the other dimension associated with cognition or, as it was named, prudence, whose characteristics are linked to consumer's rationale, which were not used in this study, since impulsivity relates to emotional attributes.This assumption gains strength when Puri (1996) identifies consumers with rational or prudent characteristics show an inverse behavior to those whose traits are emotional or hedonic (impulsive) because their actions are the result of the evaluation of costs and not of the offered benefits.Given this, we can consolidate the proposition that the promotion of monetary sales and specifically the discount, reinforced by a high attractiveness, essentially works the rational aspects and prudence at the expense of emotional variables of lack of control, which are characteristic of impulsivity formation (Puri, 1996).Thus, consumers can view this offer as something negative, especially about the evaluation of the benefit of the good offered.It is important to warn, however, that investigations by other means, such as the use of different scales than the used here, for example, those developed by Rook and Fisher (1995), Beatty and Ferrell (1998) or even the construction of new scales would be interesting to ratify or not these results.
At the management level, the results suggest that a more attractive promotion is not positively related to the impulsive characteristic of the consumer and its intention to purchase a particular product.So one might think that, for this public, it is enough to have a promotion, even if it is not attractive, in which the profitability return margin tends to be higher, and it will relate positively with the purchase intent.Thus, the results suggest that administrators can perform not so attractive campaigns in order to have an adequate rate of return and increase its margin of profitability with the impulsive public.One has to think also the focus of each business segment and its customers, who must demonstrate different behaviors.
In H 5b hypothesis, it was expected that the attractiveness of the promotion would positively moderate the relationship of the hedonic perception of the product and the purchase intent of the discounted product.The results demonstrate that the difference between the groups was significant, observing the effect of moderation; however, they were the opposite of the admitted relationships by hypothesis.Again, as described above, and aiming to increase understanding of the phenomena, we made the procedure suggested by Aiken and West (1999), shown in Figure 4.

Fernando de Oliveira Santini, Cláudio Hoffmann Sampaio, Marcelo Gattermann Perin and Valter Afonso Vieira
The results showed that when attractiveness is high, hedonism results on the purchase intent are higher; however, when attractiveness is low, with one standard deviation below the average, the hedonism results on purchase intent are smaller.Therefore, we observe, at first, that the behaviors found in moderation of attractiveness and the impulsivity and hedonic perception constructs of the product are quite close.As noted, for both behaviors, the high attractiveness did not moderated positively the relationship with the intention of buying a discounted product.This suggests and reinforces the previous reflection, concerning the non-confirmation of the hypothesis that the perceived attractiveness of a discount may have much more rational relationships and utility nature such as, for example, the feeling of prudence, than aspects associated with hedonic peculiarities such as, fun, excitement, pleasure and personal gratification (Babin et al., 1994;Childers, Carr, Peck, & Carson., 2001;Hirschman & Holbrook, 1982) whose feelings are quite close to impulsive characteristics (Dholakia, 2000;Piron, 1991;Puri, 1996).
The results reinforce also an assumption of previous researches, such as those carried out by Inman et al. (1990), demonstrating consumer responses to insignificant actions of discounts in relation to hedonic products offered.Observing this assumption, echoes the cognitive strategic strength that, through the psychoanalysis vision, consists of individual defense mechanisms (Davidoff, 1983), making it interesting to relate the example described by the author who says, "while we feel safe and in the midst of abundance, we can imagine catastrophic scenes and feel miserable" (Davidoff, 1983, p. 560).Thus, it is feasible to suggest that too much discount brings a discomfort state to the consumer, strengthening a sense of caution highlighted earlier, relating negatively with the hedonic characteristics.Moreover, we may suggest that a high discount rate may disqualify one of the main characteristics of hedonic products, which is the exclusivity associated with the good (Hirschman & Holbrook, 1982;Spangenberg et al., 1997) and may be, moreover, a significant obstacle in the way of the formation of pleasure and excitement needs, essential in the formation of hedonic consumption (Dhar & Wertenbroch, 2000;Ramanathan & Menon, 2006).Finally, the results presented in this research allow us to suggest that the high degree of discount weakens the hedonic behavior dimensions highlighted by Arnold and Reynolds (2003), such as the feeling of belonging to certain groups and adherence to innovative styles.This contribution makes us rethink that success or failure of a promotion is not necessarily on the attractiveness of benefit, as previously guided in some studies (D'Astous & Landreville, 2003;Liao, 2006;Simonson et al., 1994), but in relation to the perception of the product and the company's business segment.
Therefore, it is interesting to think on the managerial possibilities arising from the findings.Managers who work with products or services whose answers are more hedonic than utilitarian by their clients must think carefully before making a discount promotion, especially if the percentage of benefit is large, because facilitating the acquisition of a product essentially hedonic, caused in this case by the discount promotion, may have a reverse reaction in their customers, since the charm of being able to obtain this product becomes commonplace to be reached and, thus, can relate much more to utilities benefits than hedonic ones.The results also suggest that it may be appropriate to use small percentage discounts for the arousal of purchase, always taking care that the benefit does not reach the point of destroying the hedonic perception of the product.This becomes interesting for administrators, mainly those working with hedonic goods, to carry out campaigns whose margin of profitability should not be much changed in relation to the usual ones.Therefore, it should be more adequate to perform a non-monetary nature promotion than a monetary one, as already suggested by Chandon et al. (2000).
Finally, H 5c hypothesis provided a positive moderating discount attractiveness between perceived financial risk and purchase intent of the promotion.Unlike previous cases, in this case, no moderation ratio was detected, since the difference in behavior between the groups was quite similar.
A first assumption that one can raise about non-moderation of discount attractiveness in the relationship between financial risk and purchase intent of discounted product is that, in this study, it was decided to evaluate only the financial dimension of the construct.Accordingly, maybe the fact of significantly decrease the original value of a product (37%) might have stimulated other elements such as, for example, psychological risk and social risk (Jacoby & Kaplan, 1972), which were not treated here.This assumption is consistent with the suggestions made in the Garretson et al. (1999) study, when they say that the perception of quality linked to a product or service should always come first and promotional tools are secondary.Thus, it can be suggested that the discount attractiveness was not enough to change the financial risk behavior, because, even with no significant differences, there was a greater negative perception in financial risk ratio and purchase intent for the high attractiveness group (β = -0.217)and lower for the low attractiveness group (β = -0.191).Thus, the above propositions are reinforced, that the high attractiveness of a discounted product can stimulate the sense of prudence, rather linked to the cognitive dimension of the consumer, and thus accelerate a sense of financial risk perceived in relation to a good.
Analyzing from the perspective of academic contribution, the findings open up alternative sources for greater investigation of this matter, not only by the size of financial risk, but also by the feelings linked to it, such as psychological risk, social risk and also the perception of quality linked to the good.In management terms, the results point to interesting possibilities and alternatives to reduce a perception linked to an offered product.For example, the results suggest that discount actions are not the right ones to stimulate the purchase of a product whose perception is of financial risk.Perhaps more institutional or disclosure actions such as, for example, advertising

AN ANALYSIS OF THE INFLUENCE OF DISCOUNT SALES PROMOTION IN CONSUMER BUYING INTENT AND THE MODERATING EFFECTS OF ATTRACTIVENESS
campaigns, can be more effective.Another assumption to think about is that perhaps the non-monetary promotions such as gifts and promotional distribution of prizes can be more effective than the discount campaigns, considering that, without altering the original price of a product, this can be measured without changing the quality perception of it, or other related feelings.

CONCLUSIONS AND FUTURE REFERRAL
This study aimed to analyze consumer behavior in a discount promotion, analyzing positive or negative relationships that could explain such behavior and evaluating the moderating effect of perceived attractiveness in this process.The theme "sales promotion" and its investigation still have a wide field for debate because, despite being a widely used and considered technique in the management field, it is a subject little discussed in academia (Alvarez & Casielles, 2005;D'Astous & Landreville, 2003;Jones, 2008).To this end, the study proposed here sought to contribute to a better understanding of behaviors related to sales promotion and its relationship with the consumer's purchase intention.
The study presents methodological limitations, detected during the work, which will be addressed below, in the same way that new research subjects for the deepening possibilities will be suggested.The first limitations that can be highlighted concern the sampling technique, the audience to which the survey was conducted and the characteristic of the experiment used.The composition of the sample of students from a single course and college, despite having been justified, may cause some bias in the results.Further research may use samples with different characteristics, to ratify or not the results found in this study.It would also be interesting to consider the possibility of companies to do this research to their customers.The sampling technique used, with non-probability, do not allow to assign the results across the board within the studied population.Thus, one may suggest the application of the research with probabilistic sampling techniques, allowing generalization for the sample used.As for the experimental characteristic used, even with the justifications presented in the methodology section, it is assumed that it has limitations to confirm the external validity of the research.For this reason, it is important to suggest and conduct a research that include field experiments to ratify or not the results detected here.
In relation to the research of the relationship of impulsivity in the worked context, the scale used is a limiting factor (Puri, 2006), since the average values were low, suggesting the obtainment of social adequate responses, whose behaviors are common in people with impulsive traits (Rook & Fisher, 1995).As a suggestion, we propose the application of other types of scale, perhaps linked to an impulsive purchase behavior and not to personal characteristics, in order to verify if the present results are consolidated or not.Similarly, we can test the study with different scales than the ones used to measure hedonic perception.In this line, we also suggest conducting a research with measurements characterizing the hedonic consumption rather than product features.Also in the field of hedonic perception, we suggest to make the assessment within the raised relations, utilitarian perception of the product or a possible acquisition; it would be interesting to assess whether this relationship would be stronger or weaker when compared to the hedonic perspective.
In terms of research on risk perception, the limitation is linked to the fact that the analysis was made on a single dimension (financial risk).It is suggested that, in future studies, other dimensions of risk within the alleged relationship can be evaluated.Similarly, and as noted above, the interaction between products and services and their relationships with the sales promotion techniques and perceived risk could be worked out to check for possible differences or not in this behavior.In this dimension and based on these results, it would be also valuable to analyze and compare effects not only of sales promotion, but also of advertising actions in these constructs.
It is also recommended to conduct studies to evaluate different levels of attractiveness of the discount promotion, in order to identify the point in which this action takes on a negative level of moderation.We also suggest to conduct qualitative studies in the search for a better understanding of the relationships here unconfirmed.In addition, we indicate to carry out new experiments in which different products are manipulated or even goods that stimulate an extreme behavior, for example, financial risk or hedonic perception.Finally, we suggest the expansion of this work to other areas, for example, psychology, psychiatry and sociology, in search of a better understanding of the phenomena detected.
Despite the limitations highlighted, we expect that this work contributes to a better and deeper understanding of the influences of sales promotions on consumer behavior, encouraging new theoretical discussions on the subject, as well as contribute with companies' managers to better plan their sales promotions.

REFER-ENCES
An analysis of the influence of discount sales promotion in consumer buying intent and the moderating effects of attractiveness The purpose of this paper is to analyze the influence of discount sales promotion in the purchase intention and the moderating effects of attractiveness in the relationship between intention to purchase a discounted product and the impulsiveness, hedonic perception and financial risk.Thus, an experiment involving 613 students was conducted.The hypotheses predicted that a product with discount promotion would relate positively with impulsivity, as well as with a hedonic perception about the good offered, and negatively with the perception of financial risk associated with the product offered with discount.A positive moderation was expected of the perceived attractiveness of the announced discount promotion on the intentions of behaviors.The results confirmed the hypothesis, indicating positive effects of impulsivity and hedonic perception by purchasing the discounted products, in addition to the negative link between the intention of purchasing discounted products and the perception of a financial risk.The moderating effects were not confirmed.Final considerations conclude the work.
Keywords: marketing, consumer behavior, sales promotion, discount sales promotion, attractiveness of sales promotion.

Fernando de Oliveira Santini, Cláudio Hoffmann Sampaio, Marcelo Gattermann Perin and Valter Afonso Vieira result
is in line with previous research findings that demonstrate the positive effect of sales promotions on consumer behavior (Alvarez & Casielles

Table 2 Correlation Matrix of Constructs
Note: **.Correlation is significant at the level of 0.01.