Women Prefer Wine and Liquor While Men Prefer Beer: Using an Implicit Measures Approach to Determine Consumer Behavior

Often when we see a study examining alcohol preferences and habits in people, we cannot be certain if the results are based on what the participants want to report due to potential guilt or embarrassment, or if the results are actually based on fact.  Survey-type research often runs the risk of experiencing this type of variation, which may not actually reflect what the individual or group of individuals prefers or how they behave in real world situations.

One way to work around the survey method in order to obtain a potentially more accurate presentation of consumer preference and behavior is using what is called “implicit measures”.   What this means is that the strength of an

association (i.e. how strongly one associates or attaches oneself to a particular stimuli such as a specific type of alcohol) is inferred by the behavior of an individual as opposed to simply asking the individual how they feel about that particular stimuli.  Using implicit measures allows for capturing information that is beyond the conscious control of the individual, theoretically giving a more accurate representation of their preferences and consumption behaviors.

The short communication (i.e. quick study) presented today aimed to evaluate using implicit measures to determine how gender and drinker status (i.e. how much and how often one drinks) relates to selection stimuli (in this case, type of alcohol).  The implicit measures test used in this study is the Implicit Association Test (IAT).  This test uses picture or words to represent a single type of alcohol or picture or words to represent different types of alcohol.  According to the authors, this type of methodology could provide more accurate answers and ensure that the choice made actually reflects the true behavior of the consumer compared to survey methods or other methods that may be inaccurate and inconsistent.

Note: this study is a test of methods.

Methods

300 undergraduates (136 male, 164 female) between the ages of 18 and 25 (mean = 20.47) participated in this study.  They were recruited by email, and were told that they would be participating in a research study about cognitive processes and alcohol.

• 57% identified themselves as white/Caucasian;
• 30% Asian;
• 9% multiracial;
• 4% as black/African American, American Indian/Alaska Native, Native Hawaiian/other Pacific Islander, unknown, or declined answer.

During the IAT test, classification of preference should be faster when the

pairing of the target and attribute categories match the individual’s personal associations in their memories.  Two separate IAT tests were used:  alcohol approach (approach or avoid) and alcohol excitement (excite or depress).  A higher IAT score indicates a stronger relationship between alcohol and approach than alcohol and avoid, and between alcohol and excite than alcohol and depress.  In other words, a higher score indicates the individual would consume the alcohol and not avoid drinking it, and that the individual is excited about drinking a particular type of alcohol and not indifferent or “depressed” about drinking another type.

During the test, participants selected four images of alcoholic beverages out of 15.  Each image contained 3 different examples of the alcohol they were representing (i.e. 3 different types of beer or 3 different types of wine, etc).  They were asked to choose the images that corresponded to the type of alcohol they drink most often, and if they were classified as “non-drinkers”, they would select the image that corresponded to the type of alcohol that was offered to them most often.

Quantity of consumption at one time and frequency over the past 30 days were also measured.

Results

• Women chose more wine and liquor than men.
• Men chose more beer than women.
• Those considered heavy episodic drinkers (i.e. binge drinkers) chose more beer than those that were not binge drinkers.
• Binge drinkers chose less wine than those that were not binge drinkers.
• Binge drinkers chose more liquor than those that were not binge drinkers.
• Female non-drinkers were more likely to choose iced malt beverages than female binge drinkers.
• Those who consumed higher amounts of alcohol had higher IAT scores than other consumers
  • In other words, they associated more closely with “alcohol and approach” than “alcohol and avoid” and more closely with “alcohol and excite” than “alcohol and depress”.
  • Non-drinkers had significantly lower IAT scores than binge drinkers and those consuming alcohol more moderately.

Conclusions

In general, the results of this study are consistent with other studies examining the drinking habits of college students.  Males preferred more beer than females, and females preferred more wine and/or liquor than males.  In regards to the preference of those who consumed a heavy amount of alcohol at one time, those participants preferred more beer and liquor, and not wine.  Finally, females that do not drink alcohol and those that do not drink heavily at one time preferred iced malt beverages more than the other participants.

Results from the IAT test suggest that alcohol associations (i.e. “approach or

avoid” or “excite or depress”) are sensitive to differences in the amount of alcohol consumed and not to the particular type of beverage selected.  In other words, IAT test scores were significantly higher for those that consumed heavily during a drinking episode than for those that do not consume heavily at one time or that don’t drink at all.

The authors suggest that due to these results, research examining alcohol preferences among individuals or groups should use the Implicit Association Test, which would help decrease variability associated with survey-type methods that can be complicated by lying or stretching of the truth due to guilt or embarrassment.

One problem with this study is that it only includes undergraduate students; therefore the results may or may not reflect what the entire population as a whole represents.  The study also did not include other types of alcohol such as alcoholic energy drinks, even though these types of beverages are popular among college-aged students.

Ultimately, the authors claimed that the results of this study indicate that implicit measures may be a more appropriate and more accurate methodology for measure actual preferences of alcohol consumers than traditional survey methods.

I would like to have seen this study coupled with a survey method, to compare the results from the survey directly to the results of the implicit measures test.  The authors say that the implicit measures method would provide more accurate results in regards to actual alcohol consumption behavior, however, they do not describe or compare what the participants responses would have been if they were just answering the questions directly on a survey.  If they did do this, it wasn’t made clear in the paper.

What do you all think about using implicit measures methods for determining consumer behavior?  Have you ever filled out a survey asking for alcohol consumption habits?  Did you stretch the truth a little, or were you completely honest?  Please feel free to leave your comments!

Source: Lindgren, K.P., Westgate, E.C., Kilmer, J.R., Kaysen, D., and Teachman, B.A. 2012. Pick your poison: Stimuli selection in alcohol-related implicit measures. Addictive Behaviors 37: 990-993.