Stimulus-Response Variables in New Product Research

P. S. Raju, The Pennsylvania State University
ABSTRACT - This paper focuses on stimulus variables, response variables, and interrelationships between them in relation to new product research. Results from a study correlating certain specific stimulus variables and response variables are presented. Four stimulus variables, one cognitive response variable, and four behavioral response variables were included in the study. Implications of the study are discussed and issues for future research are identified.
[ to cite ]:
P. S. Raju (1979) ,"Stimulus-Response Variables in New Product Research", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 200-205.

Advances in Consumer Research Volume 6, 1979      Pages 200-205


P. S. Raju, The Pennsylvania State University


This paper focuses on stimulus variables, response variables, and interrelationships between them in relation to new product research. Results from a study correlating certain specific stimulus variables and response variables are presented. Four stimulus variables, one cognitive response variable, and four behavioral response variables were included in the study. Implications of the study are discussed and issues for future research are identified.


New product research is an important area for most companies today and a substantial amount of the marketing literature has been devoted to this area. Often quoted statistics with regard to new product failures serve to emphasize the importance of understanding consumers' responses to new products. Most available evidence in the area evolves from the "diffusion of innovations" tradition. However, on reviewing the literature in this area it is evident that in-depth studies of consumers' responses to new products and factors influencing these responses have rarely been conducted. Recently, research in the area of "variety seeking" has offered further insights into factors that influence response to new products (Venkatesan, 1972; Raju, 1977). Theories in this area have identified several useful variables. Unfortunately rigorous empirical testing of these theories has proved difficult thus far.

The need to systematically examine factors that influence response to new products cannot be denied. The present paper is a first step in that direction. More specifically, the present paper has the following objectives: 1) to offer insights on stimulus variables, response variables, and their interrelationships in new product research, and 2) to identify issues that could be addressed by future studies in the area.


The stimulus-response (S-R) model offers the simplest way of conceptualizing consumer reaction to new products. In this model response is a function of stimulus characteristics. Individual characteristics will be incorporated only in so far as the stimulus variables are consumer perceptions of the stimulus rather than absolute measures. Specific individual characteristics (such as personality, life style, etc.) and other factors that moderate the stimulus-response relationship will not be explicitly considered in this paper. Interest, therefore, will center primarily on stimulus variables, response variables, and their interrelationships.

New products can be characterized in terms of several stimulus variables. The "diffusion of innovations" literature has identified factors which influence responses to new products such as relative advantage, compatibility complexity, divisibility, and communicability (Robertson, 1972). Other variables such as novelty, ambiguity, incongruity, and surprisingness are recognized as being important in the "variety seeking" literature (Berlyne, 1960; Raju, 1977). In addition, marketing researchers have identified some stimulus variables, the most noteworthy of which is perceived risk (Bauer, 1960; Roselius, 1971).

Consumer's responses to new products can be manifested in different ways. An exhaustive classification is not possible but one could classify responses basically into two categories: "cognitive" and "behavioral." In this paper cognitive response is considered to be an overall emotional reaction toward the new product and is operationalized as relative preference for the new product over existing alternatives. Three types of behavioral response are considered: investigatory response, information seeking response, and trial response. Investigatory response pertains to exposing oneself more to a new product by looking, moving toward, or examining the new stimulus. Such responses which are "exploratory" in nature have been incorporated in the "variety-seeking" literature (Berlyne, 1960; Raju, 1977). Information seeking response comprises of acts performed to obtain more information from personal or impersonal sources. The former relates to word of mouth communication with friends, relatives, salespeople, etc., whereas the latter relates to information seeking from the product package, advertisements, and mass media. Finally, the trial response incorporates purchase of the new product. The three behavioral responses identified could be performed by a consumer to first investigate a product, then seek information about it, and ultimately try it if she is pleased with the results of her previous responses. However, a strict adherence to such a hierarchy in all new product situations would not be expected.

The stimulus variables and response variables in a new product situation are summarized in Figure A. A study that was designed to address some probing questions regarding interrelationships between these variables will now be discussed. Other important issues in the area will also be briefly reviewed later in this paper.





The study was designed to examine a few specific stimulus and responses variables in new product situations and interrelationships between these variables. The domain of consumer packaged goods and small appliances was chosen for the study. Four stimulus variables were included, namely, novelty, incompatibility, performance risk, and social risk. Response variables included were relative preference, investigatory response, information seeking from personal sources, information seeking from impersonal sources, and trial. Specific questions addressed in the study were: 1) Are the stimulus variables correlated with each other or do they represent independent dimensions? 2) Are the response variables correlated with each other or do they represent independent dimensions? 3) Are particular stimulus variables correlated with particular response variables? 4) Do certain stimulus variables play a major role in determining response to new products? Only aggregate relationships in the form of correlations have been emphasized in this study. In that respect it should be considered an exploratory study since no attempt has been made to test specific forms of relationships among the variables. However, some helpful conceptual ideas for this purpose have been offered on the basis of the results from this study.


Respondents in the study were forty women selected on the basis of a convenience sample, whose cooperation was obtained through women's' clubs in the area. Monetary incentive was provided to the clubs to encourage participation by their members. Respondents ranged from 22 to 67 years of age with the median age being 34 years. They were almost equally divided between the categories of full time employment, part-time employment, and full time homemakers. Finally, ninety percent had graduated from high school, with about thirty percent having a college degree. The sample is thus somewhat biased toward the more educated consumer but well balanced in terms of age and employment.


Data on stimulus variables and response variables for twenty new products were collected by means of a questionnaire. The twenty new products were carefully selected. Several past issues of "Advertising Age" were monitored for announcements regarding new product introductions in the market and brief descriptions of these new products were also obtained. The new product alternatives generated were then screened and twenty products selected for inclusion in the study. The criteria used in screening the alternatives were, 1) the products should be of interest to most women, 2) a respondent should be able to decide by herself fairly quickly how she is likely to respond to the new product (products for which joint decision making was likely were therefore eliminated), 3) they should be consistent in terms of the general category to which they belong so that wide variations in price and other extraneous factors could be avoided, and 4) they should either not have been introduced in the market or should have been in the market only for a short period at the time of the study. The twenty new products selected on the basis of these criteria are listed with their descriptions in the Appendix.

Measurement of Stimulus Variables. A brief description of each new product was provided to respondents in the questionnaire. The descriptions were designed to provide just sufficient information to enable respondents to Judge what the product would be like and how it was intended to function. After each new product description, measures of four stimulus variables were obtained. The four stimulus variables were novelty, incompatibility, performance risk, and social risk and these were operationalized in terms of respondents' perceptions of the new products. Other stimulus variables were considered but eliminated either because they were not relevant to the products considered or did not lend themselves to operationalization with product descriptions. Operationalization of the four stimulus variables is summarized in Table 1.



Measurement of Response Variables. One cognitive response Variable and four behavioral response variables were also measured for each product description. The cognitive response variable was relative preference and the behavioral response variables were investigatory response, information seeking from personal sources, information seeking from impersonal sources, and trial. The cognitive response variable was measured after each product description. For measuring the behavioral responses, each response was described in turn and the likelihood of performing that response for each of the twenty new products was obtained. Operationalization of the response variables is shown in Table 2.


Pearson product moment correlations were obtained between relevant variables for each product separately. Since perceptions and responses were of interest it was considered desirable to include only respondents who were in the potential market for each product in the analysis for that product. The information on whether each respondent could or could not use each of the twenty new products was obtained in the questionnaire for this purpose. Tables 3 to 6 show the correlations between different sets of variables. In each case the table provides the number and percentage of products that showed significant correlations at the 0.10 probability level as well as the range of these significant correlations. Each of these tables is briefly discussed below.

Table 3 shows correlations between the four stimulus variables. There were no consistently strong correlations between particular stimulus variables. A significant correlation between incomparability and perceived risk (both performance and social) was shown by a fair number of products. The lack of strong correlations between the stimulus variables is, perhaps, indicative of the fact that these variables represent independent dimensions of a stimulus. Thus, if a product rates high on one dimension such as novelty it need not also rate high on another dimension such as incompatibility.



Table 4 shows correlations of the cognitive response variable with stimulus variables and behavioral response variables. Among the stimulus variables incompatibility and novelty showed stronger relationships with relative preference, the correlations being significant for 40% and 35% of the products respectively. Incompatibility was correlated negatively with relative preference as would be expected. However, novelty was correlated positively with relative preference indicating that for the type of products considered, consumers prefer those that are more different from existing alternatives. Perceived risk, and especially performance risk was correlated with relative preference for few products. This may be due to the nature of products included in the study which were all consumer packages goods and small appliances. Perceived risk probably determines preference at the brand level for these products rather than at the product level. With respect to response variables, relative preference was more strongly related with trial and investigatory response. Sixty-five percent of the products showed a significant correlation between relative preference and trial and forty percent showed a significant correlation between relative preference and investigatory response. Information seeking from personal sources and from impersonal sources showed significant correlations with relative preference only for 15% of the products.





Correlations between response variables are shown in Table 5. These variables were significantly correlated between themselves for a fairly large number of products. Except for the correlation between information seeking from personal sources and from impersonal sources which was significant only for 35% of the products, all other pairs of variables showed significant correlations for at least 65% of the products. These results suggest that different behavioral responses may be dependent on each other in some systematic manner and that the performance of one response could lead to other responses. However, the weaker relationship between the two types of information seeking responses suggests that each may be relatively independent of the other even though both are related to other responses. This could happen, for example, if consumers prefer to seek information either from personal sources or from impersonal sources but rarely do both.



Table 6 shows correlations between stimulus variables and behavioral response variables. Considering each behavioral response variable separately it can be seen that the stimulus variable most related to it is incompatibility. In the case of the trial response, which is of most interest to marketers, incompatibility and novelty were the two most important variables, with perceived risk showing insignificant correlations for practically all products. It is surprising that although performance risk were significantly correlated with other response variables for a few products, those correlations were all negative. One would generally expect investigatory and information seeking responses to increase with perceived risk. A possible explanation for this is offered latter.



A summary of the results is as follows:

1. The stimulus variables are unrelated for most of the products considered and seem to represent independent dimensions.

2. Among the stimulus variables, incompatibility and novelty are most related to relative preference. Incompatibility is negatively correlated with relative preference and novelty is positively correlated.

3. Among the behavioral response variables, investigatory response and trial are most related to relative preference. Relative preference and trial show a strong relationship with significant correlations for 65% of the products considered.

4. Among the stimulus variables, incompatibility is most correlated with each of the behavioral response variables. Novelty also seems to have some influence on trial. Performance risk and social risk have no influence on trial but show negative correlations with other response variables for a few products.

5. The behavioral response variables show significant correlations between themselves for several products. However, information seeking from personal sources and from impersonal sources seem to be relatively independent although each is related to other behavioral response variables.

The next section will be devoted to analyzing the above results in terms of implications for marketing researchers.


One major issue for further investigation is whether stimulus variables identified in the literature represent independent dimensions as suggested by this study for some of the variables. Future studies should include more stimulus variables such as complexity, incongruity, perceived conflict, etc. One problem with these variables is the difficulty in operationalizing them. Research effort should therefore be directed to operationalization problems also. In essence, the area of new product perception requires more research attention.

The correlations of performance risk and social risk with relative preference and behavioral response variables are interesting. Firstly, for products included in this study, perceived risk has very little influence on relative preference and behavioral response variables. This seems to point to the fact that risk by itself is not of much concern to consumers in these product categories. There are probably other important variables that determine their responses to new products. Secondly, for a few products perceived risk is negatively correlated with investigatory and information seeking responses. This seems to indicate that in cases where risk is important, consumer are passive with respect to this variable. The tendency of consumer is, therefore, to ignore the product rather than investigate it or seek information about it. If this is true it has important implications for marketing products, because the burden of reducing perceived risk falls completely on the marketer. The conclusion, however, may not be valid for products where all alternatives have about the same amount of risk or for product categories that consumers have great need for. Thus, in buying a car for the first time a consumer would be expected to investigate alternatives and actively seek information about them.

The stimulus variables incompatibility and novelty are also interesting in several ways. Since these variables are to some degree positively correlated (Table 3) but each is related in the opposite manner to relative preference, the implication for the marketer is that a new product should be designed to be optimally different from existing alternatives to obtain positive responses from consumers. The nature of the relationships between novelty, incompatibility, relative preference, and trial are particularly interesting. Both incomparability and novelty are related to relative preference which in turn is correlated with trial. Thus, these stimulus variables probably influence behavioral response by means of the cognitive response. In the case of novelty this notion is reinforced because its relationship with relative preference (Table 4) is somewhat stronger than its relationship with trial (Table 6). However, in the case of incompatibility its direct relationship with trial is somewhat stronger than its relationship with relative preference. This raises the possibility that incompatibility might influence willingness to try a product directly without the intermediate stage of relative preference. In generalizing this idea to several stimulus variables one could say that stimulus variables can influence willingness to try a product in two ways, either directly or through intermediate cognitive responses. The two ways are not mutually exclusive so that a particular variable may sometimes operate both ways.



Considering the relationship between cognitive response and behavioral response it can be seen from Table 4 that relative preference has the strongest relationship with trial. Table 5 shows that several of the response variables are also correlated among themselves for a fairly large number of products. This suggests the hypothesis that relative preference first influences the interest in trying a product which then influences other types of behavioral responses. This hypothesis is worthy of further consideration because it runs contrary to the general notion that a consumer, if he prefers a new product, will first investigate it, then seek information about it, and finally decide to try it.

The implications discussed above indicate that there are several issues that remain to be investigated in the area of new products. At the heart of these issues is the question of what causal linkages exist between stimulus variables, cognitive response variables, and behavioral response variables. One possible model of causal linkages between variables is shown in Figure B. The model shows the influence of two major stimulus variables, incomparability and novelty, and treats them as independent dimensions. Both these stimulus variables influence intention to try the new product by means of the cognitive response variable, relative preference. However, incompatibility also has a direct effect on intention to try the product. The intention to try the product in turn influences the intention to seek information about the pro product. Information seeking from personal sources and from impersonal sources are shown as being independent of each other. The intention to seek information then influences the intention to investigate the product. The correlations obtained in this study seem to support this type of causal linkage. Other such causal models can also be conceptualized. These causal models can be evaluated by means of path analysis (Blalock, 1974; Van De Geer, 1971) for various products and a general causal model could be developed. One interesting observation can be made from the model in Figure B with respect to causal ordering of the variables. The model indicated that the consumer first determines her intention to buy the product and then decides to seek information or investigate the product. However, when the actual behaviors are performed the order is likely to be reversed, so that the results from investigating or information seeking will determine whether the consumer will try to product or not. The causal ordering of the behavioral intention variables, which are represented in Figure B and were included in the study, is therefore the reverse of the order in which consumers are likely to actually manifest these behaviors.

In summary, stimulus and response variables in relation to new products deserve more attention from marketers and marketing researchers. This paper has examined interrelationships between certain specific stimulus and response variables. Implications for future research have also been outlined. The major areas for further research are identification and classification of stimulus and response variables into meaningful categories, operationalization of variables, and identifying causal relationships between variables. Finally, one could extend the S-R type model and consider the moderating influences of individual characteristics and other important factors.



Instant Yogurt Drink: This is a new lowfat yogurt drink which tastes like a rich, creamy milkshake. It comes in four flavors: strawberry, peach, blueberry, and raspberry. The drink is packaged in 10 oz. cartons, which you only shake and pour to prepare.

Sandwich Pockets: These are flat circles of white bread which, when cut open, have pockets in the center. They can be used to make pizza, garlic bread, or a sandwich that doesn't fall apart.

Dry Soup Helper: This is a dry soup to which you add the meat. It comes in five varieties, and contains vegetables, stock, and seasonings.

Angled Toothbrush: This is a toothbrush which is constructed for most effective teeth cleaning. It has the softer bristles recommended by dentists, and a longer handle which is angled like a dental instrument, designed to reach back teeth effectively.

Multi-Surface Beautifier: This liquid restores and beautifies many surfaces, including vinyl, rubber, leather, and wood. It comes in 8 oz. plastic pump spray bottles.

Chocolate Mike Concentrate: This liquid milk additive comes in a plastic squeeze bottle and can be used to make chocolate milk. It dissolves easily in hot or cold milk.

Wine by the Glass: This wine is packaged in 6.3 oz. bottles with a plastic drinking glass shrink-wrapped to it. It comes in four varieties--Chablis, Vin Rose, and Burgundy.

Boil in Bag Rice: This rice is packaged in individual plastic cooking bags. It should be boiled right in the bag and is guaranteed to cook perfectly every time. There is no need for refrigeration and it can be stored on the shelf. Cooking time is 15 minutes, but it can remain in hot water for an additional 30 minutes without becoming gummy.

Tea for Automatic Coffee Machines: This tea is especially blended for use in automatic coffee machines. It has an extra fine texture so you get a deeper brew for making iced tea. This tea can be made in minutes in automatic machines.

Liquid Non-Scratch Cleanser: This is a liquid household cleanser with a mild abrasive. The cleaner removes dirt like a scouring powder but it does not scratch. It can be used to clean many surfaces.

Combination Cake & Cookie Mix: This product is a powdered mix which comes in a box and can be used to make either cake or cookies with a slight variation in the recipe.

Detergent with Fabric Softener: This detergent has fabric softener right in it, so there is no need to add the softener during the rise cycle or in the dryer.

Shelf Top Sealed Entrees: These entrees come packaged in a single servings in a foil pouch. To prepare, you simply boil them for five minutes, and remove from the pouch. There is no need for preservatives, and the pouches can be kept indefinitely on the shelf without refrigeration.

Concentrated Gravy and Sauce Bars: These are refrigerated gravy and sauce bars which can be heated to form the basis for any sauce or gravy. They come in three flavors: Beef, Chicken, and White Sauce. You Just add the bar to meat drippings for home-made gravy.

Fried Chicken Maker: This is a new appliance which prepares fried chicken that tastes like chicken from a carry-out restaurant. The appliance is a pressure fryer which prepares chicken in about 15 minutes and comes in 4 or 6 quart sizes.

Low Calorie Frozen Cakes: These frozen cakes have one-third fewer calories than others and contain no artificial sweeteners. They come in three flavors: yellow with chocolate icing, chocolate with chocolate icing, banana.

Turnable Teeth Comb: This comb has nickel-silver teeth which are designed to rotate through and de-tangle wet or dry hair. The comb also massages the scalp.

Toaster Fries: These are frozen potato sticks similar to french fries. They come in square sheets of eight sticks which are cooked in the toaster and then broken apart to form potato sticks.

Champagne Like Soft Drink: A soft drink with the taste and effervescence of champagne. It has no alcohol content.

Coffee & Grain Instant Beverage: This is an instant coffee and grain beverage, made of coffee, wheat, and molasses. It is designed to look and taste like coffee.


Bauer, R. A. "Consumer Behavior as Risk Taking." In R. S. Hancock (Ed.), Dynamic Marketing for a Changing World. Chicago, Illinois: American Marketing Association, 1960.

Berlyne, D. E. Conflict, Arousal, and Curiosity, New York: McGraw-Mill, 1960.

Blalock, H. M., Jr. Causal Inferences in Nonexperimental Research, University of North Carolina Press, 1964.

Raju, P. S. "Theoretical Perspectives on Exploratory Behavior: A Review and Examination of Their Relevance for Consumer Research," Paper #67, Working Series in Marketing Research, The Pennsylvania State University, October, 1977.

Robertson, T. S. Consumer Behavior, Glenview, Illinois: Scott, Foresman and Co., 1972.

Roselius, T. "Consumer Rankings of Risk Reduction Methods," Journal of Marketing, Vol. 35 (January 1971), pp. 56-61.

Venkatesan, M. "Cognitive Consistency and Novelty Seeking." In S. Ward and T. S. Robertson (Eds.), Consumer Behavior: Theoretical Sources, Englewood Cliffs, New Jersey: Prentice-Hall, 1973.

Van De Geer, J. P. Introduction to Multivariate Analysis for Social Sciences, San Francisco, California: W. H. Freeman and Co., 1971.