The consumption of convenience foods: reference groups and eating situations

Abstract

The theory of reasoned action was applied to study situational influence on the consumption of TV dinners. We investigated five situations, which were either time-related (weekdays vs weekends) or social (“dinner alone”, “dinner with family”, and “dinner with friends”). The intention to use a TV dinner decreased from “alone” via “with family” to “with friends”, but did not differ between weekdays and weekends. Subjective norms were a stronger influence on intentions than attitudes in all situations, except for “weekdays” and “dinner alone”. Primary reference groups were a stronger influence on intentions than secondary reference groups, and the motivation to comply with a particular reference group increased substantially when it joined for dinner. Consumption frequency for TV dinners was higher in households where the person responsible for meal preparation held a paid job, and it was positively related to the number of hours that this person was employed.

Introduction

The role of convenience in food choice and preference has been described in several studies. It has been shown that perceived convenience is strongly correlated with preferences for meals and snacks (Rappoport, Peters, Downey, McCann & Huff-Corzine, 1993), and with preferences for different packages for beverages (van Dam & van Trijp, 1994). Also, Anderson and Shugan (1991) found that the increasing demand for chicken meat is partly due to the increasing appreciation of the fact that chicken is easier to prepare than beef. The importance of convenience reflects a tendency to minimise the time and effort that is put into the preparation of meals. Various strategies may be followed to accomplish this. Examples include eating convenience foods, buying take-away meals, eating in (fast food) restaurants and using convenient appliances such as microwaves (cf. Bryant, 1988, Darian, and Tucci, 1992, Kim, 1989, Oropesa, 1993, Reilly, 1982). These strategies, however, will not be applied in all situations, and different strategies are applied in different situations (Darian, and Tucci, 1992, Nickols and Fox, 1983).

Research on food choice and preference has paid little attention to situational influence (Meiselman, 1996). The most important studies are those of Lahteenmaki, and Tuorila, 1998, Miller and Ginter, 1979, Termorshuizen, Meulenberg, and Wierenga, 1986. Miller and Ginter found that eating situations had a strong impact on the choice for fast-food restaurants. Termorshuizen et al. found that the attitude toward a beverage could vary considerably between situations (i.e. breakfast, lunch, afternoon and evening). For example, attitudes toward coffee were highest for evening and lowest for breakfast, while the reverse was true for fruit yoghurt drinks. Lahteenmaki and Tuorila found similar effects for the intention to consume milk drinks and fruit juice.

We seek to expand extant knowledge on situational influence in the consumption of convenience foods. Specifically, we will study the role of eating situations in the consumption of TV dinners. TV dinners are meals bought in stores, and prepared at home by re-heating. The consumption of such meals reduces the time and effort necessary for meal preparation. As such, TV dinners are a prototypical convenience food. Because the use of TV dinners seems to be more appropriate in some situations than in others, they provide an interesting subject for studying situational influence.

Consumer researchers have found that consumer behaviour is better understood when situational influences are taken into consideration (cf. Assael, 1995). Predictions of consumer choice through first and second order Markov models have been shown to improve significantly when situational influences are accounted for (Umesh & Cote, 1988). Miller and Ginter (1979) found that the performance of multi-attribute models improves when a situational variable is incorporated. Like Lahteenmaki and Tuorila (1998), Miller and Ginter found that behavioural intentions vary across situations. In addition, they found that the evaluations of several attributes of fast-food restaurants (i.e. “speed of service”, “convenience”, and “variety of menu”) differed significantly between situations. Similar results were reported by Termorshuizen et al. (1986) for their study on beverages. We now can formulate the following two hypotheses:

H1a

Consumers' intention to consume TV dinners varies significantly between eating situations.

H1b

The evaluations of (at least some) attributes of TV dinners differ significantly between eating situations.

In line with Miller and Ginter (1979), and Lahteenmaki and Tuorila (1998), we investigate situational influence in an experiment-like setting, in which situational variables are manipulated within subjects. Specifically, we ask respondents to answer questions about their preference for convenience foods in different situations. The explicit measurement of preferences in clearly defined situations avoids confounding effects of situational variables. To study the role of situations in the consumption of convenience foods in more detail, we focus on two aspects of particular interest, namely, time and social surroundings. Below, we will discuss these factors, and their relevance for convenience food consumption.

Convenience foods are “...fully or partially prepared foods in which significant preparation time, culinary skills, or energy inputs have been transferred from the home kitchen to the food processor and distributor” (Capps, Tedford & Havlicek, 1985). According to this definition, time saving is an important characteristic for convenience foods, which has been acknowledged by several studies on the effect of time pressure on the consumption of convenience foods (cf. Darian & Tucci, 1992). In these studies, time pressure is measured by the working status of the person who is responsible for meal preparation. The results of these studies have been unequivocal. The use of convenient appliances like dishwashers and microwaves is higher when the person who is responsible for meal preparation holds a job. The same finding has been reported for the consumption of meals away from home, but not for the consumption of convenience foods (Bryant, 1988, Darian, and Tucci, 1992, Kim, 1989, Yen, 1993). Darian and Tucci attribute these results to the fact that meals away from home are a suitable alternative for cooking, while convenience foods are not. Previous research on the relation between convenience food consumption and working status was done in the US. The study of this relation in a different cultural setting, namely, the Netherlands, provides an opportunity to examine whether these findings generalise across cultures. Regarding the relation between working status and the consumption of convenience foods, we hypothesise that:

H2a

The consumption frequency for TV dinners is higher for households in which the person who is responsible for meal preparation has a paid job.

H2b

The consumption frequency for TV dinners increases with the number of hours of paid employment of the person who is responsible for meal preparation.

It is widely recognised that consumer behaviour is influenced by social surroundings (Bearden, and Etzel, 1991, Belk, 1975). For food products, however, it is often argued that the role of social influence is negligible, and that behaviour is primarily determined by individual preferences and nutritional needs (Bearden & Etzel, 1991). Some authors have opposed this view. Furst, Connors, Bisogni, Sobal and Winter Falk (1996), for example, found that the management of (family) relationships is an important influence on food choice. Pliner and Chaiken (1990) and Rozin (1996) point out that food choice is strongly dependent on the social and cultural context of the choice situation. These authors argue that food preferences vary widely between cultures and social classes, and that the acceptability of food items is shaped by education, upbringing, advertisements, press reports and word-of-mouth. A more immediate form of social influence is provided by the direct social context of eating. Meals are often eaten in the presence of others, and the acceptability of alternatives depends on the company for which the meal is prepared (cf. Furst et al., 1996). Obviously, the social influence on the intention to consume a TV dinner when a consumer is alone, differs widely from the social influence on this intention when the consumer has a family dinner or a cosy meal with friends. We hence hypothesise that:

H3a

Social setting has a strong impact on the social influence on consumers' behavioural intentions regarding the consumption of TV dinners. In particular, social influence is strongest when others are part of the eating situation.

Social influence is performed by reference groups, consisting of one or more persons or institutions. Reference groups can be classified as primary or secondary. Primary reference groups are characterised by strong (emotional) ties and frequent interpersonal interactions. This category generally includes partner and children. Secondary reference groups may include friends, colleagues, and institutions like government and media. Emotional ties with these groups are weaker and interactions are less frequent. In general, primary reference groups are more likely to influence behaviour than secondary reference groups (Assael, 1995). However, as we have mentioned earlier, the influence of a particular reference group depends in part on the social context of the meal. When friends or family join for dinner, for example, the person who is preparing the meal will be more likely to comply with their preferences and opinions.

H3b

When social surroundings are not specified, primary reference groups exert a stronger influence on the consumption of TV dinners than secondary reference groups do.

H3c

The extent to which a particular reference group influences behaviour is affected by the social surroundings of the behaviour. In particular, the influence of a reference group increases when this group (or person) joins for dinner.

We studied the influence of situational variables within the well-known framework of the theory of reasoned action (TRA, Fishbein & Ajzen, 1975). TRA is widely adopted by social scientists, and provides a reasonably adequate forecast of consumer behaviour and intentions (cf. Ajzen and Fishbein, 1980, Shepherd, 1990). TRA is based on the assumption that an individual's behaviour is largely determined by his or her behavioural intentions. These intentions (BI_act), are determined by the person's attitude towards the behaviour (A_act), and subjective norms (SN_act):

$$\text{BI}{}_{\mathrm{<mtext>act</mtext>}}\text{=β}{}_1\text{A}{}_{\mathrm{<mtext>act</mtext>}}\text{+β}{}_2\text{SN}{}_{\mathrm{act}}$$

To predict buying behaviour, a consumer's attitude towards buying a product is inferred from the summed expectations with regards to n relevant product attributes. These expectations are obtained by multiplying attribute beliefs b_i with their evaluations e_i:

$$\text{A}{}_{\mathrm{<mtext>act</mtext>}}\text{=}\text{∑}\text{i=}\text{l}\text{n}\text{b}{}_{\mathrm{i}}\text{×e}{}_{\mathrm{i}}$$

The subjective norm is assumed to be made up of the social influence exerted by m relevant reference groups. This influence is equal to the sum of normative beliefs NB_j, that is, the individual's estimate of the opinion of reference group j, weighted by MC_j, the extent to which the individual is motivated to comply with these beliefs:

$$\text{SN}{}_{\mathrm{<mtext>act</mtext>}}\text{=}\text{∑}\text{j=}\text{l}\text{m}\text{NB}{}_{\mathrm{j}}\text{×MC}{}_{\mathrm{j}}$$