Adrienne Keller, Ph.D., Research Director; Jennifer Bauerle, Ph.D., Director; James Turner, M.D., Executive Director. National Institute for Social Norms, University of Virginia, Charlottesville, VA 22903, February 2008/Updated January 2011.
Download the original PDF (10 pages, 381 KB)
Contents
- Introduction
- The hypothesized causal pathway
- Logic of Social Norms Interventions
- Necessary Evidence to Demonstrate Effectiveness & What’s Included in the NCHA IIa
- Overview of Methodology
- Evaluation Tool Assessing the Strength of Necessary Evidence and Inferences:
- Some Key References
National Institute for Social Norms
University of Virginia
Charlottesville, VA 22903
Authors:
Adrienne Keller, Ph.D., Research Director
Jennifer Bauerle, Ph.D., Director
James Turner, M.D., Executive Director
February 2008/Updated January 2011
Introduction
The materials that follow were developed at NSNI as a working tool primarily to assist colleges and universities who are conducting social norms marketing interventions to reduce the adverse consequences associated with high risk alcohol consumption. Since many colleges and universities participate in the National College Health Assessment (NCHA II), this document uses questions in the NCHA II as examples of available relevant information.
The model and the assessment tool, although both written in terms of social norms marketing campaigns directed towards alcohol consumption, can be adapted for other issues. Please contact us at NSNI if you would like assistance in adapting this material for other social norms marketing interventions.
The hypothesized causal pathway
The social norms model hypothesizes that, against a background of a wide variety of influences (Box 1), a person’s perceptions of peers’ behaviors and beliefs surrounding alcohol (Box 2) will have a strong influence on both personal drinking patterns (Box 3a) and use of protective strategies (Box 3b). These will, in turn, strongly influence the number, type and severity of adverse consequences associated with hazardous drinking (Box 4).
Causal Pathway
- BOX 1: Demographic characteristics (e.g., age, sex, race/ethnicity)* + Other influences (e.g., genetic, biological, familial, religious, past experiences, current socialization)
- BOX 2: Perceptions of peers’ behaviors (descriptive norms* & injunctive norms)
- BOX 3a: Personal drinking patterns*
- BOX 3b: Use of protective strategies*
- BOX 4: Adverse Consequences associated with hazardous drinking*
- Box 1 → Box 2
- Box 1 → Box 3a
- Box 1 → Box 3b
- Box 2 → Box 3a
- Box 2 → Box 3b
- Box 3a ↔ Box 3b
- Box 3a / Box 3b → Box 4
*Characteristics in bold are measured by variables in the National College Health Assessment (NCHA II) survey conducted of the American College Health Association (http://www.acha-NCHA II.org/ )
A social norms marketing intervention aims to correct the misperception of peers’ behavior in order to influence personal choices. The messages typically focus on the reality of peers’ drinking but may also include the use of protective behaviors by peers. For social norms marketing interventions, the causal pathway can be used to create a logic model that relates the intervention to goals, outcomes and impact.
Logic of Social Norms Interventions
- Underlying problems
- Personal hazardous drinking pattern
- Misperception of normative behavior
- Intervention strategy
- Social norms marketing campaign
- Intervention goal
- Correct misperception of descriptive and/or injunctive norms
- Anticipated outcomes
- Increase use of protective behaviors*
- Decrease hazardous drinking patterns
- Desired impact
- Harm Reduction: Decrease alcohol-related adverse consequences
- Personal hazardous drinking pattern ↔ Misperception of normative behavior
- Underlying problems → Intervention strategy → Intervention goal → Anticipated outcomes (each box) → Desired impact
- Increase use of protective behaviors* ↔ Decrease hazardous drinking patterns
Evidence Needed
- Underlying problems: Prevalence of hazardous drinking patterns**; prevalence of misperceptions**; association between the two**
- Intervention strategy: Campaign conformed to social norms principles & reached intended audience
- Intervention goal: Prevalence and/or degree of misperception decreased**
- Anticipated outcomes: Prevalence & number of protective behaviors used increased** and hazardous drinking decreased**
- Desired impact: Prevalence and/or number of adverse consequences decreased**
*If norms for protective behaviors were included in the marketing campaign
**Baseline assessment needed before intervention & follow-up assessment after intervention
Necessary Evidence to Demonstrate Effectiveness & What’s Included in the NCHA IIa
Sample Description: (NCHA II questions 46, 47, 51 – 55, 58, 59)
Evidence related to underlying problems
- Perception of norms related to drinking (NCHA II questions 9e, 12, 17b)
- Drinking patterns (NCHA II questions 8e, 10, 11, 13, calculated eBACb and NIAAA “binge” drinkingc)
Evidence related to intervention strategy (not included in NCHA II)
- Fidelity: Marketing materials conform to social norms principles in development strategy & content
- Development strategy involved members of the intended audience in a meaningful way
- Materials address correcting a significant misperception related to peers’ behaviors and conform to PIE principles: Positive, Inclusive, Empowering
- Reach: Marketing reaches intended audience
- “Exposure”: evidence that a majority of the intended audience was exposed to the campaign
- “Saturation”: evidence that information was seen a sufficient number of times by intended audience
- Suggested question to measure exposure: In the past [time frame], how many times have you seen [or heard] the [campaign logo/brand or message or specific campaign material]?
Evidence related to intervention goal
- There is a significant improvement from pre-intervention to post-intervention in measure(s) of normative perceptions:
- Descriptive (NCHA II questions 9e, 12, 17b)
- Injunctive (not included in NCHA II)
- Example of injunctive norm: Most [Anywhere U] students disapprove of drinking to get drunk.
Evidence related to anticipated outcomes
- There is a significant improvement from pre-intervention to post-intervention in measure(s) of protective behaviors (may not happen if campaign did not focus on protective behaviors):
- Personal protective behaviors (NCHA II question 15a through 15k)
- Prevalence
- Frequency of use
- Protective behaviors for others (not included in NCHA II)
- Prevalence
- Frequency of use
- Personal protective behaviors (NCHA II question 15a through 15k)
- There is a significant improvement from pre-intervention to post-intervention in measure(s) of patterns of drinking; for example:
- Number of drinks consumed last time partied or socialized (NCHA II question 10)
- Drinks per hour (NCHA II question 10 divided by NCHA II question 11)
- Decreased eBACb
- Number of days in previous month drank alcohol (NCHA II question 8e)
- Decreased “binge” drinkingc
Evidence related to desired impact
- There is a significant improvement from pre-intervention to post-intervention in measure(s) of adverse consequences associated with drinking (NCHA II question 16a through 16i: 9 adverse consequences)
Optional additional helpful evidence
- Level of drinking and negative consequences did NOT increase from pre-intervention to post-intervention among low risk drinkers
- Perceptions, drinking and negative consequences all changed in the desired direction from pre-intervention to post-intervention among high risk drinkers
aNCHA II is the second, and latest, version of the American College Health Association’s National College Health Assessment. A sample NCHA II survey is available at http://www.acha-ncha.org/docs/ACHA-NCHAII_sample.pdf
bAlgorithm for estimating Blood Alcohol Content (eBAC) (see Hustad, J.T.P., and Carey, K.B. Using calculations to estimate blood alcohol concentrations for naturally occurring drinking episodes: A validity study. Journal of Studies on Alcohol, 66, 130-138, 2005.):
BAC = [(consumption/2) x (GC/weight)] – (0.017 x hours)],
where consumption = number of standard drinks in a drinking session (NCHA II question 10)
hours = the number of hours over which drinking occurred (NCHA II question 11)
weight = weight in pounds (NCHA II question 50)
GC = gender constant, (9.0 for women and 7.5 for men; gender is NCHA II question 47)
cNIAAA definition of binge drinking (see http://pubs.niaaa.nih.gov/publications/newsletter/winter2004/newsletter_number3.pdf):
- BAC equal to or greater than .08. “For the typical adult, this pattern corresponds to…”
- 4 or more drinks in about 2 hours for women; 5 or more drinks in about 2 hours for men
Overview of Methodology
- Conduct a (stratified) random sample survey of intended population
- Ensure sufficient sample size for sufficient power to examine subgroups (e.g., by sex or class level) taking into account likely response rate (response rate should be at least 40%)
- Clean the data (eliminate respondents with incomplete essential data or artifactual data)
- Balance sample demographic characteristics to match population characteristics, for example:
- Sex
- Academic class level (in school populations)
- Race/ethnicity
- Establish the baseline rate of behaviors and attitudes of interest:
- Quantity/frequency of drinking: a variety of possible variables including drinks per week, drinks per weekend, drinks last time partied/socialized, eBAC
- Perception of drinking of others: perception questions must correspond exactly to quantity/frequency questions
- Use of protective behaviors (may be omitted)
- Adverse consequences associated with drinking
- Test if there is a misperception of the true drinking norm: are the perceived quantity/frequency measures significantly greater than the actual quantity/frequency measures?
- Test if this misperception is associated with hazardous drinking
- Those who drink more are more likely to perceive the drinking norm to be greater than the true drinking norm, [and close to or greater than what they themselves drink, per Berkowitz]
- Establish the percent of the population that have normative misperceptions and hazardous drinking
- Design a social norms marketing campaign
- Use representatives from intended population to design message and marketing strategies
- Ensure messages conform to PIE principles: Positive, Inclusive, Empowering
- Use a variety of messages and/or message delivery strategies
- Test if the marketing campaign reached the intended audience
- Exposure: What percentage of the population saw the messages?
- Saturation: How frequently did members of the intended population see the message?
- Conduct a follow-up (stratified) random sample survey of the intended population:
- Same questions about drinking behaviors and attitudes, protective behaviors and adverse consequences as in the first step
- Closely following conclusion of campaign
- Clean and balance sample data as in the first step
- Test if observed changes conformed to theory:
- Is there a decrease in perception of drinking norms?
- Is there a decrease in actual drinking?
- Is there an increase in protective behaviors (if included as focus of campaign)?
- Is there a decrease in adverse consequences?
Evaluation Tool Assessing the Strength of Necessary Evidence and Inferences:
This tool is designed to help you assess the strengths of the six essential components of the social norms marketing campaign. Remember that the logic connecting your intervention to observed changes is only as strong as your weakest component. Ideally, each area should score at least 8; when any area scores less than 6 that is an indication that improvements in that area would strengthen your evidence of effectiveness.
| 1. The data are valid and reliable: | Completely | Mostly | Somewhat | Not at all or information not available |
|---|---|---|---|---|
| a. The sampling strategy was unbiased and yielded at least a 40% response rate. | 3 | 2 | 1 | 0 |
| b. The resultant sample matched the population in key demographic/descriptive variables (or was weighted to match the population). | 3 | 2 | 1 | 0 |
| c. All variables of interest were assessed using standardized measures with demonstrated validity and reliability in the population of interest. | 3 | 2 | 1 | 0 |
| Total Score for Data Quality: | ||||
| 2. Hazardous drinking and misperceptions of norms are underlying problems: | Agreement among all measures | Agreement among most measures | Agreement among some but not most measures | Not at all or information not available |
|---|---|---|---|---|
| a. Hazardous drinking patterns are prevalent. | 3 | 2 | 1 | 0 |
| b. Misperceptions of drinking norms are prevalent. | 3 | 2 | 1 | 0 |
| c. There is a positive correlation between hazardous drinking patterns and misperceptions. | 3 | 2 | 1 | 0 |
| Total Score for Evidence of Underlying Problems: | ||||
| 3. The intervention was well implemented: | Completely | Mostly | Somewhat | Not at all or information not available |
|---|---|---|---|---|
| a. The social norms campaign conformed to social norms marketing principles. | 3 | 2 | 1 | 0 |
| b. The campaign reached at least 80% of the intended audience. | 3 | 2 | 1 | 0 |
| c. Over 50% of the intended audience saw campaign message(s) at least 5 times. | 3 | 2 | 1 | 0 |
| Total Score for Evidence of Well-Implemented Intervention: | ||||
| 4. Intervention Goal*: Following the campaign, accuracy of perceptions increased: | Change is statistically significant and clearly meaningful | Change is small but statistically significant | Change occurred but is not statistically significant | No evidence, no change, or change in wrong direction |
|---|---|---|---|---|
| a. The perceptions of norms that were a focus of the campaign were more accurate. | 3 | 2 | 1 | 0 |
| b. The perception of norms that were not a focus of the campaign did not improve in accuracy. | 3 | 2 | 1 | 0 |
| c. High-risk populations (defined by drinking behavior and/or risk factors) increased accuracy of perceptions. | 3 | 2 | 1 | 0 |
| Total Score for Evidence that Intervention Goal was Achieved: | ||||
| 5. Anticipated Outcomes*: Following the campaign, decreased risk for harm was observed: | Change is statistically significant and clearly meaningful | Change is small but statistically significant | Change occurred but is not statistically significant | No evidence, no change, or change in wrong direction |
|---|---|---|---|---|
| a. Hazardous drinking behaviors decreased overall . | 3 | 2 | 1 | 0 |
| b. eBAC decreased overall. | 3 | 2 | 1 | 0 |
| c. High-risk populations experienced decreased risk for harm. | 3 | 2 | 1 | 0 |
| Total Score for Evidence that Intervention Outcomes were Achieved: | ||||
| 6. Desired Impact*: Following the campaign, significant harm reduction was observed: | Change is statistically significant and clearly meaningful | Change is small but statistically significant | Change occurred but is not statistically significant | No evidence, no change, or change in wrong direction |
|---|---|---|---|---|
| a. There is a decrease in adverse consequences associated with drinking. | 3 | 2 | 1 | 0 |
| b. There is an increase in percent experiencing no adverse consequences. | 3 | 2 | 1 | 0 |
| c. High-risk populations experienced fewer adverse consequences. | 3 | 2 | 1 | 0 |
| Total Score for Evidence that the Desired Impact was Achieved: | ||||
| Overall Total Score for Necessary Evidence & Inferences: | ||||
*For multi-year projects, the measures of perceptions, behaviors and consequences following each year of intervention become the baseline measures for the next year.
Some Key References
Bauerle, J (2003). The University of Virginia’s Social Norms Marketing Campaign. The Report on Social Norms: Working Paper #11. Little Falls, NJ: PaperClip Communications.
Bauerle J, Burwell C & Turner JC (2002). Social Norms Marketing at the University of Virginia. Presented at the Annual Meeting of the American College Health Association, May 23, Washington, DC.
Berkowitz AD (2002). Responding to the Critics: Answers to Common Questions and Concerns About the Social Norms Approach. The Report on Social Norms: Working Paper #7. Little Falls, NJ: PaperClip Communications.
Borsari B & Carey KB (2003). Descriptive and injunctive norms in college drinking: a meta-analytic integration. Journal of Studies on Alcohol, 64(3), 331-41.
Bruce S, Hawkins P, Sharp M & Keller A (2006). Experience-Based Suggestions for Achieving a High Survey Response Rate. Report on Social Norms, http://www.mostofus.org/newsletter/index.php?newsletterID=5.
Bruce S & Keller A (2007). Applying Social Norms Theory within Affiliation Groups: Promising Interventions for High-Risk Drinking, NASPA Jl 44(1), 101-107.
Haines MP, Perkins HW, Rice RM & Barker G (2005). A Guide to Marketing Social Normsfor Health Promotion in Schools and Communities, http://www.socialnormsresources.org/pdf/Guidebook.pdf.
Keller, A. & Bauerle, J. Using a Logic Model to Relate the Strategic to the Tactical in Program Planning and Evaluation: An Illustration Based on Social Norms Interventions. American Journal of Health Promotion, 24:89-92, 2009.
Killos, L., Hancock, L., Wattenmaker, A., & Keller, A. Do ”clicker” educational sessions enhance the effectiveness of a social norms marketing campaign? Jl of American College Health, in press, 2011.
Mattern JL & Neighbors C (2004). Social norms campaigns: examining the relationship between changes in perceived norms and changes in drinking levels. Journal of Studies on Alcohol, 65(4),489-93.
Perkins HW (2002). Social norms and the prevention of alcohol misuse in collegiate contexts. Journal of Studies on Alcohol, Supplement (14), 164-72.
Perkins HW & Berkowitz AD (1986). Perceiving the Community Norms of Alcohol Use Among Students: Some Research Implications for Campus Alcohol Education Programming. International Journal of the Addictions, 21(9/10), 961-976.
Turner JC, Bauerle J & Shu J (2004). Estimated blood alcohol concentration correlation with self-reported negative consequences among college students using alcohol. Journal of Studies on Alcohol, 741-749.
Download the original PDF (10 pages, 381 KB)