Only $35.99/year

Marketing Research

Terms in this set (136)

History: external events that occurred during the experiment may influence DV

Ex. Bonobos (clothing company) found that their sales of shorts—one of their top-selling products—has decreased from June 2017 to the present. Given that Wal-Mart acquired their company in June 2017, they suspect that their association with Wal-Mart is what caused this decline.

Maturation: change in participants during experiment can influence DV

Ex. Starbucks conducts an experiment testing whether caffeine has negative effects on memory in older (80+) people. They take a baseline measure of memory, then give the participants coffee everyday for a year, and then measure memory again a year later.

Instrument variation: change in measurement instrument (e.g., researcher) may explain changes in DV

Ex. In an experiment testing how a new ad campaign influences brand attitudes, Trader Joe's measures attitudes using a 1 (dislike) - 7 (like) scale, then launches the campaign, and then measures attitudes using a 1 (negative) - 7 (positive) scale after launching the campaign


Selection bias: systematic differences between the treatment and control group explain differences in DV

Ex. Equinox gym wants to know how different types of music affect how long people spend exercising—they play rap music in the morning, no music in the afternoon, and electronic music in the evening and track how long gym visitors stay.

Mortality: drop-out of certain participants in course of experiment

Ex. To test how using social media affects life satisfaction, adults are randomly assigned to spend 0 or 30 minutes on social media everyday for a month. 1 week into the experiment, most millennials in the 0-min condition quit. The researcher finds that the 0-min condition is more satisfied with life at the end of the month.

Testing effects: simply taking part in experiment can have its own effect on DV

Ex. Duolingo wants to know if their app actually helps people learn foreign languages. They conduct an experiment where participants take a German vocabulary test, use the app for three days, and then take the vocabulary test again.

Ex. A meditation workshop wants to understand how effective their program is at reducing stress. The director measures participants' stress levels before the workshop and after the workshop and finds substantial improvement.

Regression to the mean: extreme behaviors naturally move toward the average (i.e., become less extreme) over time, which can be mistaken for effect of IV

Ex. Steph Curry selects a sample of kids who are very bad at free throws (make 5% on average) to participate in his basketball clinic. By the end of the program, their free throw percentage has improved to 15% on average, and they attribute this improvement to Steph Curry's instructions.
Population specification error
- incorrect definition of the population (i.e., sampled the wrong population)

Ex. Kellogg's conducted a survey using a nationally representative sample of parents in order to get feedback about the packaging of a new children's cereal they're developing. When the product is finally released, sales were much lower than expected.

Ex. Sephora (the cosmetics store) is trying to determine what types of sales promotions will be most effective for attracting millennials. To explore this, they sent an email survey to a random sample of students at co-ed colleges.

Ex. Vans (shoe brand) wants to evaluate their brand image among their consumers and conducts interviews with a random sample of people ages 30 through 60.

Ex. Blackberry is trying to understand why their former customers switched to different brands, so they conducted personal surveys with a random sample of current iPhone owners

Ex. Dell is trying to create an ad campaign that will attract new customers, so they contact a sample of their most loyal customers to provide feedback on the different versions of the ads.

Sampling frame error
- correct population but sampling frame is inaccurate or incomplete

Ex. Lyft conducted a survey to learn about consumers' fears regarding autonomous vehicles. To find respondents, they used a list of registered car owners provided by the DMV in various states.

Ex. Furniture retailer Design Within Reach is trying to determine what new brands to carry in their brick and mortar stores and sent a survey to a random sample of the last 1,000 customers who purchased anything from their website.

Ex. Boston College is trying to understand what motivates alumni to donate, so they got a list of alumni who had donated more than $50,000 in the past 5 years and sent a survey to a random sample of these donors.

Selection error
- biased process of selecting participants to be in sample
- Can occur even when population and frame are accurate

Ex. Bank of America is conducting interviews to understand the spending habits of adults in the US. Customers who visit bank branches on weekday afternoons are asked if they would like to participate.

Ex. The owners of the TD Garden are considering redesigning the arena. They got the contact information of all people who had purchased Bruins or Celtics tickets in the past decade and sent a survey to club seat season tickets holders to find out what they like and dislike.

Ex. The Microsoft Store in the Prudential Center is trying to understand why their store traffic is so low, so they conduct a "mall intercept" study where they ask shoppers who are walking by their store to complete a short survey.

Ex. TaskRabbit wants to understand the types of people who sign up to run errands for others, so they sent a survey to New Yorkers who had worked for them in the past three months.