June 18, 2024

The object of market research is finding out information which is usually concerning the opinions of people about something or someone. It is almost always impossible to survey (ask) everyone who may matter, say all possible voters. In this example, all possible voters would comprise the “universe.” Instead, market researchers survey some of the universe. That some is called the “sample.”

REMEMBER: The sample is the key ingredient in the statistical recipe or process. No matter how advanced the analysis, without a sample that accurately reflects the universe you have GIGO.

Perhaps nothing illustrates this point better than the famous Literary Digest poll. Literary Digest was a successful American magazine. In 1936, it published a poll predicting the Republican candidate for President, Alf Landon, would easily defeat incumbent Franklin Delano Roosevelt’s (FDR) bid to be re-elected President. The Literary Digest did use the proper analysis. But it did not consider that by only surveying (sampling) its readers plus people on readily available lists such as telephone directories, it was not sampling typical American voters.

The reason is that during the depths of the Great Depression only wealthier Americans, who tended to be Republican, could afford a magazine subscription and a telephone. Concurrently and independently of Literary digest, an generally unheard of statistician was sampling a considerably smaller number of truly representative American voters (5,000 v. 2,000,000). He accurately predicted FDR’s landslide victory. Soon after the election, Literary Digest went out of business while George Gallup went on to become perhaps the best known of all pollsters.

Literary Digest’s sample was what is called “biased.” Gallup’s sample was what is called “random.” The laws of probability dictate that a random sample will better represent the universe than any other kind of sample. However, to be random, each member of the universe must have an equal chance (equal odds) of being sampled.

Now you can see where Literary Digest had a major problem and the result was GIGO.

The first step in this random sampling process is defining the universe. For example, in politics is the universe everyone eligible to vote, registered voters, or some other group of people. A lot depends on the objective of the survey. Another point is that many surveys purport to use a random sample or make readers think the sample was random, when in fact the sample was not. That is one of the tricky things about interpreting survey results and often takes some skilled questioning to determine.

Similarly, if you are trying to determine the market potential for a new product, who is likely–not who you want–to be in your universe? Those are the people you want to randomly sample.

The only improvement in random sampling is when with high certainty, the universe can be divided into different “strata.” Randomly sampling each strata produces stratified random sample. But remember knowing the strata is critical to making stratified random samples successful predictors.

Sometimes you don’t know your universe is. One way of trying to determine that universe is with focus groups randomly drawn form the population in general. A future article will discuss focus groups.