Consumer Behavior Experiment Design
Overview and Methods
This experiment is designed and will be carried out using a randomized controlled “real” experiment for consumers who have placed orders for McDonald’s food products within the period of the study and at the same time able to access the usual company’s quarterly e-newsletter. A group of 2000 customers will be selected randomly and divided into four groups with each of the groups A, B, C, and D having 500 participants. The random selection will be done in two specific stages with an aim of eliminating systematic sampling bias. Before the experiment, the consumers will be informed about their selection and participation in the research concerning the company’s purchase systems and return conditions. This information will be passed through the company’s newsletter, which will have the details about the condition for participation.
One of the conditions is that the consumers (participants) must have used the company’s website for shopping within the period of experiment. Contrary to other experiments done in the laboratories, this is a field experiment, which means that the consumers will not to be asked to deviate from their normal consumption behavior, but use their consent in all the purchase and return programs. The 2000 consumers will participate in the experiment by purchasing the available food items with a possible return of items not exceeding 840. The data for the experiment will then be analyzed using the various types of regression analysis and the results presented in a table format.
- How does packaging styles and marketing strategies affect consumer’s willingness to successfully purchase McDonald’s food products?
- How does packaging styles and marketing strategies affect the value of orders purchased by a consumer?
- How does packaging styles and marketing strategies affect consumer’s decision to return McDonald’s food products after purchase?
- How does packaging styles and marketing strategies affect the value of items returned by the consumer after purchase?
Like in all other known transactions, customers’ purchase process is affected by factors such as quantity, quality, packaging styles, marketing strategies and prices of the products. For this experiment, the focus will be on how factors like food packaging and marketing strategies affect the decision by consumers to purchase and consume McDonald’s food products. The experiment is based on the assumption that consumers are already aware of market prices and may not bother much about changes in future prices, but rather are derived by the packaging and marketing strategies used by the company. This means that the decision by the consumers to successfully purchase and consume the food products or return the food products is triggered by the quality of packaging and the marketing styles the company uses. In this experiment, packaging styles and marketing strategies for the food products will be assigned letters FQ and FD respectively and their impact on the sales volume, returned items and respective values analyzed using both the logit and normal regression models.
In order to determine successful purchases and return behaviors of consumers, the experiment will include use the number of processed order (purchase), the value of purchased items, the number of returned items and the value of the returned items as the dependent variables. The number of processed purchase and the number of food items returned by the customer will be tested using the logit regression model. This is because with the logit model, it will be possible for the researchers to estimate the probabilities associated with each of the dependent variables. The measures will be presented in terms of probabilities or proportions to allow for future predictions about the changes in consumption behavior.
On the other hand, the value of purchase items and the value of the returned items will be analyzed using the normal regression model because the coefficients will be useful in predicting the monetary values related to each variable. With the monetary values, the company will be in a position to determine the changes in consumer preferences as seen by the corresponding utilities and levels of satisfaction. The monetary values will represent the kind of satisfaction consumers get from the company’s product, which will prompt them to purchase more of the products and return zero products. That is to say, a higher monetary value spent on the food products will indicate higher levels of satisfaction with lower monetary values will be an indication of diminishing levels of satisfaction. In order to measure the mentioned four dependent variables, the experiment design will use four separate regression models as shown below.
The logit model for the analysis of the probability of the products purchase by the consumers:
Logit (Pu) = b0 + b1FD + b2FR + (FD*FR)b3 + e
The logit mode for the analysis of the probability of the products returned by the consumers:
Logit (Re) = b0 + b1FD + b2FR + (FD*FR)b3 + e
The normal regression model for the value of products purchased by the consumers:
Log (PuIteValue) = b0 + b1FD + b2FR + (FD*FR)b3 + e
The normal regression model for the value of products returned by the customers:
Log (ReItemValue) = b0 + b1FD + b2FR + (FD*FR)b3 + e
Based on the experiment overview, methods and variables, the design follows four important study hypothesis which relate consumers’ purchase behaviors to both quality and quantity of the food products. The hypotheses are as follows:
- Packaging styles and marketing strategies affect consumer’s willingness to successfully purchase McDonald’s food products
- Packaging styles and marketing strategies affect the value of orders purchased by a consumer
- Packaging styles and marketing strategies affect consumer’s decision to return McDonald’s food products after purchase
- Packaging styles and marketing strategies affect the value of items returned by the consumer after purchase
In general, the experiment will use ANCOVA tests to combine both the qualitative variables (packaging styles and marketing strategies) and quantitative variables (number of orders, purchase values and returned order values) to determine the relevant coefficients, b0, b1, b2, and b3.