Stratified sampling has many applications in marketing and politics. It is often used to determine the marketability of new products, design advertising campaigns or to shape political strategies by discovering the concerns of various groups within the general population. It provides the tools to pinpoint specific target groups or to appeal to a wide range of groups by isolating commonalities between divergent groups.
Selecting the Sample
A grocery store wants to survey their customers to find out if they are satisfied with the service and selection provided by the store. They have the names and information of 5000 customers and the sample should include 500 customers or 10% of the total group. The store is located in a suburban area and the customers are 85% Caucasian, 10% African American and 5% Hispanic. A computer generated simple random selection might not include representative percentages of the different types of customers.
Stratified Sample Groups
The groups may be broken down by ethnicity before the random sample is selected which will result in a stratified sampling example which includes each group with the same percentage of representation they have as customers within that group. In other words, a simple random selection may indicate that the store provides adequate selection and service to 90% of its customers while a stratified selection might show that 50% of the Hispanic customers and 20% of the African American customers were not satisfied with the store.
Marketing Implications of Stratified Sampling
Now assume that the store is located in an area where the general population is made up of 70% Caucasian, 20% African American and 10% Hispanic residents. This stratified sampling example shows the store's customers do not represent the general population of the area. Store management can take steps to adjust the store selection, improve service to the minority populations and increase the overall customer base of the business. A simple random sampling would not provide specific information about the store's client base.
Advantages over Simple Random Sampling
In a simple random sampling of customers, the smaller groups might not receive proportionate representation which could result in an analysis that failed to indicate the dissatisfaction of the smaller groups within the store's customers. Since the store already has a disproportionate percentage of Caucasian customers when compared to the general population, the opportunity to target other groups to increase the customer base might be overlooked.
Importance of Selecting Subgroups
When conducting an analysis it is important to select the correct subgroups for the most accurate and meaningful results. In this stratified sampling example, assume 10% of the Caucasian population is Jewish. This is relevant for a grocery store since Jewish customers have dietary laws and will patronize stores that cater to their specific needs. Failure to include Jewish customers as a subgroup could mean overlooking an opportunity to increase the customer base.
Stratified sampling not only provides information about the general population being studied, it also provides information about subgroups. Identification of subgroups is a key to useful analysis with this method and researchers should consider the subject of the analysis and the possible subgroups within a general population when designing studies.