IS MAKING A SHOPPING LIST AN EFFECTIVE WAY OF PREVENTING CONSUMERS FROM MAKING IMPULSE PURCHASE?

In this research an experimental study has to be designed. The goal is to develop an experimental procedure that will answer the research question of choice. Subjects for research questions were given to choose from. The research question we have chosen is: “Is making a shopping list an effective way of preventing consumers from making impulse purchases?” This research question is chosen because it seems quite logical that people with a shopping list will only buy the products on that list. They most probably will look only on the list and look for the products in the store rather than looking around in the store and pick products they see and think they’re also nice to have. To make things a bit more interesting, we have modified this research question and added an extra variable is added, being “time”. It might be interesting to see whether people who have a shorter time to shop will only buy the most necessary products. People with less time cannot walk around in the store calmly and evaluate many products for necessity. Our research question is as follows: “Are the use of a shopping list and available time for shopping predictors for impulse buying?”


Introduction
In this research an experimental study has to be designed. The goal is to develop an experimental procedure that will answer the research question of choice. Subjects for research questions were given to choose from. The research question we have chosen is: "Is making a shopping list an effective way of preventing consumers from making impulse purchases?" This research question is chosen because it seems quite logical that people with a shopping list will only buy the products on that list. They most probably will look only on the list and look for the products in the store rather than looking around in the store and pick products they see and think they're also nice to have.
To make things a bit more interesting, we have modified this research question and added an extra variable is added, being "time". It might be interesting to see whether people who have a shorter time to shop will only buy the most necessary products. People with less time cannot walk around in the store calmly and evaluate many products for necessity. Our research question is as follows: "Are the use of a shopping list and available time for shopping predictors for impulse buying?"

Variables
Within this research there are two independent variables (shopping list and time) and one dependent variable (% of impulse purchases). These variables can be put in a 2x2 design. Independent variables + levels: 1. Use of a shopping list (2 levels > shopping list and no shopping list) 2.
Time spent in the store (2 levels > large amount of store time and small amount of store time) Dependent variable 1. % unplanned purchases in total purchases (the larger the part, the more impulse purchases)

Design: 2x2
Indeed, we define 'impulse purchase' as any unplanned purchase. By doing this, we follow the definition used by POPAI 1 . The downside of this definition that it does not account for purchases that were not 'planned' (as defined by: not on the shopping list), but were still needed. Nevertheless, we assume that whether a product purchase was planned prior to entering the store is a good indicator for unplanned purchases.

Hypotheses
Given our research question, it logically follows that we have the following two hypotheses for the main effects: H1 -The use of a shopping list reduces the % of impulse purchases H2 -Reducing the amount of time spent in a store reduces the % of impulse purchases Furthermore, we expect that an interaction effect is present between the two independent variables since a customer has more chance of buying unwanted products if he spends more time in the store. The hypothesis for this interaction effect is as follows: H3 -The effect of the use of a shopping list on the % of impulse purchases is stronger when more time is spent in the store.

Manipulation of independent variables and measurement of dependent variable:
The independent variables should be manipulated in order to test if there is a significant difference between the groups (shopping list vs. no shopping list). Nevertheless, it is not valid to manipulate shopping time and the use of a shopping list since it will influence the shopping behavior of the customers. For instance, some may be used to shopping with a shopping list and some may be not. Therefore, a large group of customers (about 400) will be selected after their shopping experience in the supermarket (4 different supermarkets), and will be questioned about their use of shopping list (yes or no) and their familiarity with the supermarket (only familiar customers will be selected in the experiment). After these two questions, the customers have to assess their shopping basket, together with a researcher. The products that are bought unreflective, unplanned (in an impulse) have to be selected, and the percentage of these products of total products in the basket case have to be computed per customer. This questioning will be recorded on camera and together with the camera at the entrance the total time spent in the store can be computed per customer.
The upside of this research design is that customers will not know that they are taking part in a research until after they are done shopping (unobtrusively). They will exhibit their actual behavior. The downside is that this research will take quite some time to complete since individual customers have to be tracked to record their time spent in the store. Nevertheless we think the cost savings of not having to use a laboratory setting outweighs this downside.
This design implies that a between-groups design will be used, with four groups:

Hypothetical outcomes
The results of our results will yield some outcome percentages for each of the groups, which can be summarized into a graph as shown on the next page. Furthermore, if our hypothesis would indeed be true, we would see the interaction effect as shown in the second figure on the next page.

Research limitations
This experimental design contains a few limitations that can influence the results of the study: -Customers that are questioned about the products they purchased can be dishonest about products they bought on impulse. A logical reason for this issue is the shame that a customer can feel towards the researcher about an impulse product. -Other factors can influence impulse buying of customers. Store environment, for example, could be a predictor for impulse buying. To minimize this influence, four supermarkets are used in the experiment.