|
Field
Abstract
|
Before
For cost-minimizing consumers, optimal shopping depends on the structure of the grocery shopping environment. Specifically, consumers need to account for various components of the shopping environment, for instance, product prices (deal hunting), shopping trip frequency, inventory holding cost, and fixed cost of shopping trips. Given such complexity of grocery shopping environments, understanding consumers’ grocery shopping behavior has received considerable attention in the industrial organization and marketing literature.
Within the consumer grocery shopping literature, the topic of consumer stockpiling has been studied in the context of price deals, i.e., consumers stockpile when price deals are available, so they can maintain their consumption at minimum cost. However, stockpiling can also happen due to fear of grocery shortages. When consumers cannot find their preferred products, shopping fixed costs (i.e., opportunity costs) can increase, and consumers would like to stockpile to avoid the high fixed costs. This study aims to assess the impact of grocery shortages, via an increase in shopping fixed costs, on consumers’ grocery purchase behavior in a lab environment. Understanding the impact of grocery shortages and fixed costs on shopping behavior, and generating scientific evidence, is relevant for food product companies and grocery stores. Food companies may need to consider product availability, especially during periods of grocery shortages. Grocery stores may need to think of store policies during grocery shortages.
Our initial proposal is to conduct the following: Implement a four-arm randomized controlled trial (RCT) among university students in a lab environment, to evaluate the causal effect of grocery shortages (via high fixed costs) on average purchase quantity (per transaction). Beside the control arm, the other three arms will have high fixed cost, low price, and purchase-limit treatments, respectively. The high fixed cost and low-price treatments are predicted to increase average purchase quantity, i.e., lead to stockpiling, in comparison to control arm. The purchase-limit treatment indirectly imposes a high fixed cost; however, it is to be explored whether it leads to stockpiling or not.
Regarding average purchase quantity as an outcome, we have one main hypothesis, and two exploratory hypotheses. The main hypothesis is that average purchase quantity will increase in the presence of grocery shortages (which is implemented via high fixed cost treatment) in comparison to the control arm. The two exploratory hypotheses are that each price decrease and purchase-limit will increase average purchase quantity in comparison to the control arm.
When we compare average purchase quantity in a treatment arm against the control arm, we are essentially estimating the magnitude of stockpiling (beta coefficient), i.e., how much did average purchase quantity increase in a treatment arm in comparison to the control arm.
|
After
For cost-minimizing consumers, optimal shopping depends on the structure of the grocery shopping environment. Specifically, consumers need to account for various components of the shopping environment, for instance, product prices (deal hunting), shopping trip frequency, inventory holding cost, and fixed cost of shopping trips. Given such complexity of grocery shopping environments, understanding consumers’ grocery shopping behavior has received considerable attention in the industrial organization and marketing literature.
Within the consumer grocery shopping literature, the topic of consumer stockpiling has been studied in the context of price deals, i.e., consumers stockpile when price deals are available, so they can maintain their consumption at minimum cost. However, stockpiling can also happen due to fear of grocery shortages. When consumers cannot find their preferred products, shopping fixed costs (i.e., opportunity costs) can increase, and consumers would like to stockpile to avoid the high fixed costs. This study aims to assess the impact of grocery shortages, via an increase in shopping fixed costs, on consumers’ grocery purchase behavior in a lab environment. Understanding the impact of grocery shortages and fixed costs on shopping behavior, and generating scientific evidence, is relevant for food product companies and grocery stores. Food companies may need to consider product availability, especially during periods of grocery shortages. Grocery stores may need to think of store policies during grocery shortages.
Our initial proposal is to conduct the following: Implement a four-arm randomized controlled trial (RCT) among university students in a lab environment, to evaluate the causal effect of grocery shortages (via high fixed costs) on average purchase quantity (per transaction). Beside the control arm, the other three arms will have high fixed cost, low price, and purchase-limit treatments, respectively. The high fixed cost and low-price treatments are predicted to increase average purchase quantity, i.e., lead to stockpiling, in comparison to control arm. The purchase-limit treatment indirectly imposes a high fixed cost; however, it is to be explored whether it leads to stockpiling or not.
Regarding purchase quantity as an outcome, we have one main hypothesis, and two exploratory hypotheses. The main hypothesis is that, on average, purchase quantity will increase in the presence of grocery shortages (which is implemented via high fixed cost treatment) in comparison to the control arm. The two exploratory hypotheses are that each price decrease and purchase-limit will increase average purchase quantity in comparison to the control arm.
When we compare purchase quantity in a treatment arm against the control arm, we are essentially estimating the magnitude of stockpiling (beta coefficient), i.e., how much did average purchase quantity increase in a treatment arm in comparison to the control arm.
|
|
Field
Power calculation: Minimum Detectable Effect Size for Main Outcomes
|
Before
We determine the sample size to detect a standardized effect size of 0.3 in average purchase quantity among the treatment arms with 80% statistical power, and 5% (two-sided) level of significance. Since each participant will go through 4 treatment arms, we assume a correlation of 0.5 in average purchase quantity outcome across the 4 arms. We postulate the means of our outcome measure to be 5 (Control arm), 10 (Fixed Cost arm), 10 (Price arm), and 20 (Purchase-Limit arm). Using Stata’s power analysis program for repeated-measure analysis of variance, we estimate a sample size of 121 participants.
|
After
We determine the sample size to detect a standardized effect size of 0.3 in average purchase quantity among the treatment arms with 80% statistical power, and 5% (two-sided) level of significance. Since each participant will go through 4 treatment arms, we assume a correlation of 0.5 in average purchase quantity outcome across the 4 arms. We postulate the means of our outcome measure to be 5 (Control arm), 10 (Fixed Cost arm), 10 (Price arm), and 20 (Purchase-Limit arm).
|