The Standard Model Operational
Insert Tab A into Slot B
This chapter details the operational behaviors needed to create a “Standard Model” intervention. If you properly enact each, you will have all the resources you need to implement, manage, and evaluate a campaign. The Model is designed to create an intervention that:
Widely distributes strong arguments aimed at key beliefs that drive behavior change.
If you understand this sentence backwards, it means that this campaign will achieve behavior change by identifying beliefs that underlie behavior change through a process of elaborated argumentation that causes people to generate favorable thoughts about the new belief. To ensure that favorable thinking, you must make sure that everyone in the population gets multiple opportunities to think about those arguments and you do this by widely distributing the messages. This can be schematically represented as:
In other words, make sure everyone gets the message (Reception); thinks carefully and favorably about the arguments in the message (Processing); changes key beliefs about the behavior as a result of the argument processing (Response); and finally changes the behavior (Behavior Change).
The Standard Model takes this basic three step cascade and details each stage with a theory drawn from the literature. (You could substitute different theories for each stage to create a different model. The key element is whether the new theory successfully generates Reception, Processing, or Response.) For Reception, the Model uses the principles from mass media advertising; for Processing it uses the Elaboration Likelihood Model; and for Response it employs the Theory of Planned Behavior.
To create a Standard Model communication campaign you must take Four Action Steps
1. Define the Behaviors
2. Determine Beliefs that Drive Response
3. Create Messages that Affect Processing
4. Develop an 80% Reception Plan
The following sections will detail each action step. In particular it will describe the operational behaviors you need to enact.
Action Step 1: Define The Behaviors
Goal: Define Specific, Concrete, And Observable Behaviors
A communication intervention is more likely to succeed when the target behavior is explicitly defined on four dimensions: Who does What Where and When. If you produce bad behavior definitions or if you thoughtlessly alter behavior definitions through the various stages of the intervention, you will fail. Get this step right and everything else will be a lot easier; get this one wrong and everything will be a lot harder.
Behavior is not understood or expressed in one simple word like “vote” or “buy” or “volunteer.” Behavior must be approached as Who does What When and Where. If you cannot express the Behavior with this specificity, then your intervention will most likely fail.
Behavior change is the ultimate outcome of a persuasion intervention. If you don’t get change, nothing else matters. You failed. You didn’t do a persuasion intervention. You did an “awareness” intervention. You did an “educational” intervention. But you didn’t do a persuasion intervention and that’s what we said we were going to do. You can’t get “change” unless you know exactly what it is you are trying to change. Thus, if you do not TACTfully define your Behavior, you don’t have a good way to measure or count that thing you’re trying to change. If you do not TACTfully define your Behavior, you will have huge difficulty finding the Response variables that drive the Behavior. If you do not TACTfully define your Behavior, you will have no good way to come up with Arguments for the Processing stage. If you do not TACTfully define your Behavior, your Reception plan will be ignored by your targets because the messages will seem pointless, irrelevant, and silly to them because you didn’t define the Behavior TACTfully.
Can I drive this point into the ground? Can I go on like Aunt Tillie having a bad hair day? If you don’t TACTfully define each and every Behavior you want to change in a persuasion intervention, you will fail.
So, how do you define behavior? Get TACTful.
“Who” is the person you want to change. It is more helpful at this stage of planning to think of “who” as an individual rather than a group.
“What” is the observable concrete action you want “who” to perform. That means everyone can see the person doing “it.” The “what” should be as simple an action as you can make it. You should not use “and” or “but” or “if, then” language. Just a simple action.
“Where” and “when” essentially define the context in which “who” does the “what.” Again, aim as explicit statements. “Most of the time” is not as explicit as “ 5 or more days a week.”
Consider this as a behavior definition that could be found in the pages of applied research literature on public health: As a result of this intervention more people will live a healthy and longer life through wiser choices in diet and exercise.
Sounds great and might even get some funding, but this is a dog that won’t hunt, beg, or fetch, but just roll over and play dead. Look at all the imprecision. “Longer life” sounds good, but operationally, what is it? Ten years? Ten days? And a “wise” choice? Who’s wise here: the Center for Science in the Public Interest or the National Beer Wholesalers Association? What’s exercise? A serious hand of Texas Hold ‘Em or a walk in the woods? (In my experience, “longer” life means “a quantitative result our statistician said was statistically significant,” while “wiser” means “the invisible college of our peers and colleagues,” and “diet and exercise” means “whatever we do for recreation that you don’t.”)
Typically, an intervention focuses upon one behavior definition. If, however, the intervention seeks multiple TACTs, then this process must be executed for each one. For communication purposes, it is crucial that you isolate and separate each behavior. If only for planning purposes, you should think as if you have as many campaigns as you have behavior definitions. Thus, you need to execute each step in this manual for each targeted behavior.
Many cooks will spoil this broth. Assemble a very small group of key intervention people who will have leadership positions in the intervention. You do not want a large group here. You need focus. If you are working alone, that’s fine, too. It does help, however, if you can test your ideas with another person just to see if they pass the “laugh” test (if you tell people what you are doing and they don’t laugh, then you’re in the ballpark whereas if they laugh you’re going back to the drawing board).
On a large writing surface, like a whiteboard or flip board, write out the basic equation of Who does What When and Where, then discuss or think about the meaning of each term. After everyone gets a good sense of the terms, each team member should write down as many Who Does What When and Where statements as they think are crucial. Take five or ten minutes here. Work separately and just let everyone create. (If you are doing this one by yourself, try doing this activity several times over a couple of days.)
After everyone is done, have each person write their behavior definitions in separate areas of the whiteboard. Now, just look at them and consider what you’ve got. You should see a lot of similarity across the definitions. You should also see some interesting creative ideas. Take a little time and discuss what you’ve produced. After about 10-15 minutes of discussion, you might repeat the prior exercise. Sometimes after a little talk, people get new ideas and you can grab them now. Just take another 5 quiet minutes, let everyone write, then re-board the new thoughts. Look at everything again and start making choices.
As you start making choices keep one idea in mind: Be Obvious! Each behavior definition should be simple, clear, direct, observable, measurable, changeable, etc. Don’t get subtle, wise, or clever. You will probably come up with 5 to 10 good statements. Remember that for every behavior statement you have a “different” intervention!
Explicit behavior statements are the essence of a strong campaign. If the behavior definitions are vague, ambiguous, or confusing, your communication will necessarily be vague, ambiguous, and confusing. Using the simple schematic of Who does What When and Where you can easily create behaviors that are concrete, observable, and measurable. Remember if you want to use communication to change behavior, you must define each behavior you want to change.
Action Step 2: Determine The Beliefs That Drive Response
Goal: Construct A TpB Model For Each Behavior Definition.
Conceptually, this step looks like this:
This is just the standard Theory of Planned Behavior model. All we do here is fill in the blanks for each behavior.
1. Review the literature.
Hey, maybe somebody has already done this, published it, and all you have to do is read it. Chances are good there is Something Out There and it would be wise to read it and think about it. Don’t repeat their mistakes.
Chances are also good that there is nothing Out There that is quite like your plans. All politics is local and so is all intervention. You’ll have to do some of your own work after all.
2. Conduct Belief Elicitation.
Goal: find a representative sample of our target population and let them tell us what they naturally believe about the behavior.
Get 10 to 15 people who are currently actively engaged in the behavior and another 10 to 15 people who currently are not engaged in the behavior. We’ll label them, respectively, as Doers and NonDoers. Make sure that within the two groups of Doers and NonDoers, the participants are “different” from each other (gender, age, culture, etc.). You will now execute what is essentially a focus group method with a collection of 7-12 people with a group leader and an agenda aimed at generating interaction and comment among the participants to answer questions you’ve got. Thus, you’ll need a quiet room, a blackboard or flip chart, a table and chairs, and a leader with proven experience and success. If you have a total of 30 respondents you’ll need to run 4 or 5 groups. It is probably easier if you keep the Doers and NonDoers in different groups (Doers tend to like other Doers better than NonDoers and vice versa, so the ice breaking and group solidarity goes faster). If scheduling issues make this division difficult, don’t worry about it. Just make sure you can identify individual output as a Doer or a NonDoer.
Have a group leader who knows how to make strangers feel at ease run the groups. You can do a bunch of activities here to break the ice and build solidarity. When the group is warmed up:
1. Carefully and fully describe the target behavior to the group. Talk about the definition and make clarifications to insure that everyone understands the same thing.
2. Ask each person to write down all the things that makes is Easier or Harder to do the behavior. Allow several minutes for people to think and write. When everyone is done, lead a discussion about the Easier-Harder ideas. After the discussion dies, ask people to look at their list and make any changes they could make given the discussion.
3. Ask each person to first write down all the people who Approve or Disapprove of doing the behavior. Allow several minutes for people to think and write. When everyone is done, lead a discussion about the Approve-Disapprove ideas. After the discussion dies, ask people to look at their list and make any changes they could make given the discussion.
4. Ask each person to first write down all the things that are Good or Bad about doing the behavior. Allow several minutes for people to think and write. When everyone is done, lead a discussion about the Good-Bad ideas. After the discussion dies, ask people to look at their list and make any changes they could make given the discussion.
5. Take the three lists of Easy-Hard, Approve-Disapprove, and Good-Bad and read each statement provided by all respondents and make sure that the statement addresses one and only one idea. For example, somebody may write that “Authority figures (like my doctor or my minister) would approve of me eating 5 servings.” You should break “Authority Figure” into two categories, “My doctor” and “My Minister.” You now have a large list of all the individual beliefs your respondents generated for three categories: Attitude (Good-Bad), Norm (Approve-Disapprove) and Control/Efficacy (Easy-Hard). Read the master list of statement and remove duplicates, so that you have a Master List of unique belief statements.
Goal: Create a TpB survey and administer to large groups of Doers and NonDoers to construct a full TpB model.
This is a basic Research Methods 101 activity. Take a theory of behavior change, draw a representative sample of your target population, administer a reliable and valid survey of the theory variables, and quantitatively analyze the variables according to the model of the theory. Publish in a peer review journal, get tenure, fame, and fortune, film at 11.
Construct the Survey
1. Take the master list of beliefs generated from the Doer-NonDoer activity and format each belief statement into the desired TpB structure.
2. An Example
“Fruits and vegetables cost too much money compared to other foods.” is a belief statement (Good-Bad) generated in the focus group. This statement can be structured on the survey as,
21. “Fruits and vegetables cost too much money compared to other foods.”
Strongly Agree Agree Uncertain Disagree Strongly Disagree
1 2 3 4 5
Do this with each unique belief statement. You should have a survey with 30 to 50 items. To avoid tricky context effects, it is advisable to break the survey into the 3 major categories of Attitude, Norm, and Control with all the belief statement for each category appearing together. You do not need to provide a distinctive header in the survey for each category, just string them in line.
You also need to include at least three other key items on the survey. I’ll suggest wording in my examples, but please use your own preference. There is more argument about how to ask about eating than with virtually any other health and safety behavior I’ve encountered. I don’t care what you prefer, just be consistent with it throughout the intervention and all assessments. Don’t flip back and forth in wording preferences or you will reap the whirlwind.
Past Behavior Item.
In the past 30 days, I have eaten 5 or more servings of fruits and vegetables
Everyday Most days Some days A couple of days Never
Intention to perform Behavior Item.
In the next 30 days, I intend to eat 5 or more servings of fruits and vegetables every day of the week.
Strongly Agree Agree Uncertain Disagree Strongly Disagree
1 2 3 4 5
A Doer is someone who eats 5 or more servings of fruits and vegetables at least 5 days a week every week. A NonDoer is everybody else. Are you a Doer or a NonDoer?
You could also create this classification using combinations from the Past Behavior and Intention items or by creating your very own Stages of Change scheme.
You are creating a basic TpB survey that uses a quantitative scale to measure beliefs about Attitude (Good-Bad), Norm (Approve-Disapprove), and Control (Easy-Hard), intention to perform, and past behavior. There is a ton of variation in how people write these items. My experience indicates that you really, really, really need to measure beliefs with the specific wording supplied by the focus group respondents. Lots of TpB work also includes items related to the “evaluation” element in the classic B*E equation. Lots of TpB work also includes items at a higher level of abstraction, one’s overall Attitude, Norm, and Control rating rather than specific elements. If you want to be a TpB purist, that is fine. Just be consistent in your measurement. If you use complex measures in the formative stage, then use complex measures in the summative or evaluative stage. (And remember that you might be doing measurement over the telephone.)
3. Other Items on the Survey
You are going to execute a serious Research Methods 101 data collection with this survey. You might consider adding other variables on it for other research purposes. Some enterprising grad student might benefit by including the Next New Theory with this survey as part of a thesis or dissertation project. Maybe an enterprising professor is developing preliminary studies for a new grant application. Maybe you just really like looking at crosstabs based on gender, age, cultural tradition, political affiliation, or Star Trek convention attendance. You can include “extra” items along with the basic TpB components (Attitude, Norm, Control, Intention, Past Behavior). Please just think about the size of the survey and a rational arrangement of the questions. Make sure you’re not wearing out your respondents and that you don’t inadvertently confuse people by randomly sequencing your questions.
4. Administer the Survey (the Sample)
The gold standard here is drawing a random sample from a well defined population. Everyone knows the perils of bad sampling. Those perils are especially perilous at the formative stage. A biased sample here will destroy your chances for success even before the intervention hits the ground. Thus, you want to get as close to the gold standard as you can afford in the largest sense of the term. Without spending too much time, money, or effort, get a good large heterogeneous sample of several hundred respondents. If you end up with 78 captives from a public health grad class, you will deserve the pain, humiliation, and failure that will inevitably arise during the intervention.
You will probably have to administer the survey to many different groups at different times and places to get 250 or more total. Develop a protocol for administration, train on it, and enforce it. If you break protocol on one administration, note it, code it, and analyze that subgroup to see if they responded very differently. If they are weird, drop them from the larger sample and run another group if you need the sample size. Play this straight.
5. Analyze the Survey
Goal: Identify the key beliefs that discriminate Doers from NonDoers.
From my practical experience doing large scale interventions and from reading a lot of student projects plus all the research literature I’ve read indicates that typically just a few beliefs will discriminate Doers from NonDoers. Often times, these beliefs will cluster into just one of the TpB components. For example, in the “1% or Less” nutrition studies, we found that the key beliefs that separated Doers from NonDoers came out of Attitude (Good-Bad) especially with factors like taste and cost. Norms were irrelevant. By contrast in the “Wheeling Walks” exercise intervention, the key beliefs came out of Control, particularly with time schedule. Norms and Attitude were irrelevant. The same thing will happen with all behaviors. Different beliefs structures determine different behaviors.
You will have a lot of data with this survey. You will have 30-50 belief statements, intention, and behavior, and Doer-NonDoer status. The easiest way to start is to run a Discriminant Analysis with Doer-NonDoer as the independent variable and each belief statement as a dependent variable. You can play around with different selection strategies and see how the list of beliefs changes. Typically you will find 2 to 5 beliefs that almost always come up as statistically reliable (.05) regardless of selection strategy. And typically, most of those recurring beliefs will fall out of one major component (Attitude, Norm, or Control).
If you are in a hurry or not conceptually inclined, you can stop here. I would encourage you, however, to take some time here and try many different analytic approaches and turn the data like a diamond and consider a lot of facets. For example, TpB lends itself to structural equation modeling with a linear flow from the components of Attitude, Norm, and Control, leading to Intention which leads to Behavior. Run a big model on the whole sample, then break it out by Doer-NonDoer status. Do the two models still look the same? Try running SEM with just the statistically reliable beliefs statements, intention, and behavior. Do the whole sample, then the breakout by Doer-NonDoer. Take some time and use the data to understand what is going on here. You’ve got the best theory of behavior change and a deep, representative sample. It’s a living slice of action caught in data. Let it tell you what’s going on.
If you are working on a smaller scale, such as a class project or underfunded activity on your job, you should take this large, complex plan and adjust it to your more limited goals and resources. Please don’t get intimidated with all the details a full scale plan requires. Just understand each of the activities described here, then cut them down to your needs and resources. The key element here is to engage in systematic, theory-based planning. Even if you can past ten without taking off your shoes, you can still think carefully and plan ahead.
6. Commit to the Key Beliefs
You now have a crucial tool for creating change. You know how Doers and NonDoers are different. If you can change those key beliefs in the NonDoers, you can slowly change them into Doers. Commit to changing those key beliefs.
This is the most difficult part of the formative process. No matter how strong your data are, inevitably in the planning process, you will start thinking about other approaches. You will read a study that used something different. There will be somebody on the team that is a strong advocate for this new thing. You’ll start to remember something from an old grad class or a project done a few years ago. The creative team doing message development will come up with this incredibly clever approach that drops the key beliefs for a New New Thing.
Look, you are taking a serious scientific approach to get to this point. You are doing a lot of hard and grubby work based in standard theory and research principles. Don’t forget it. When the New New Thing comes along, make sure it has the same level of science behind it that you’ve got right now.
At this point you have detailed TpB statements that operationalize a TpB model for each behavior. You’ve got good data for publication in several forms. You could take the data and write it up in a journal for peer review publication. You can also take these TpB models and put them in reports and presentations to your funding agencies. The TpB diagrams are very intuitive and most folks grasp them very quickly. Just provide a few words that briefly define the key concepts, then show them the Power Point slide. Most importantly, you’ve got the Key Beliefs that distinguish Doers from NonDoers.
Action Step 3: Create Messages that Affect Processing
Goal: Create Persuasive Messages that Change Key Beliefs.
Conceptually this step look like this:
This is a highly modified version of the Elaboration Likelihood Model. It holds that people typically employ one of two different approaches to persuasive information. They either have high willingness and ability to think (high WATT) about persuasive information or else low willingness and ability (low WATT). High WATT thinkers want arguments – information that is crucial to the persuasion – and they want to think carefully and effortfully about that information. If the argument is strong, the thinking will generate favorable thoughts and lead to “Central Route ” change which is strong, long lasting, and drives behavior. By contrast, under low WATT conditions, the person cannot think about arguments and instead prefers cues or information that persuades without requiring much thoughtfulness. These cues can lead to “Peripheral Route” change which is weaker, short term, and less likely to drive behavior. We want that Central Route change and that means we’ve got to create strong arguments. How do you do that?
The first thing we have to do is get out of our human nature. We cannot get out of our skins, our heads, our point of view. We tend to approach persuasion with a feral craftiness, heh-heh-heh, they’ll never see this one coming! – when all we’re doing is what seems convincing to us and not to the target. If you look closely in the mirror at those moments you should see someone who looks like Wile E. Coyote.
They are two primary ways you get in the target’s head. First, you focus your attention upon them for a long time, observe them carefully, question them in hidden ways, in general you play Jane Goodall and live with the apes in their jungle for many years. In other words, you become a monkey. Now, obvious this monkey thing is metaphor and Dr. Goodall did not become a version of Tarzan or a crazy Sci-Fi X-Files character. But she lived with the apes, watched the apes, and thought a lot about what she saw. She got into the heads of monkeys so much so that she could psychologize like a chimp. So, if I was opening a new Burger King in the Gombe Reserve and wanted to get the chimp business, Jane Goodall would be my first marketing consultant.
Okay, this is cute and fun and nice and we can find the nugget of truth with a smile, but don’t get carried away with it. You can count the number of people who have truly become the monkey on one hand. Garry Trudeau, the Doonesbury cartoonist, did a biting sendup of the fake monkey in the 1970s when he portrayed a hip religious leader Reverend Scott Sloane, Jr., the fighting young priest who talks to the young. The Rev knew the famous dictum from theater, dress the part and the role plays itself. He then added a little righteous anger and as far as the Silent Majority was concerned, the Rev had become the monkey. Of course everyone over 30 thought the Rev was the cat’s meow while everyone under 30 knew he was just wearing a monkey suit.
Now, if you’re like me and you don’t like bananas (and don’t even start with me about the rich source of potassium argument) and don’t like living in the jungle or woods or anyplace without cable, plasma screens, and central air, then you might want to consider a second approach. Second, you do effective research on your target. While there is no magic bullet for persuasion, the best approach I’ve found is an empirical one that uses a guided process with empirical markers. You start off with some good ideas, then run them through the process and see what survives.
Main Point: Change the Key Belief! No matter what approach you take always remember to keep a relentless focus on changing the Key Beliefs. Your Aunt Tillie finds a particular execution to be compelling. Ohhh, it’s a perfect illustration of a fear appeal! Wow, look at the Stages of Change in that one! All foolishness. Does the message change a Key Belief? If yes, keep going. If no, try again.
Secondary Point: Everyone is a genius when it comes to messages. Everyone knows how to talk and, hey, I’ve gotten this far so I must be pretty good at communicating. Don’t listen to your gut instincts here. Use the method I outline now. Use the empirical markers. Remember, I don’t care if you think the message is cool, I care if the target population gets the message and changes key beliefs. Anything else is vanity.
Review the Literature. What message tactics did other people use? Maybe you’ll get lucky and find a published example that addressed one of your Key Beliefs. After you get through at the library, you can do your own empirical work.
1. Focus Group Message Tactics
Get 10 to 15 NonDoers and run them in a focus group method. Describe a Key Belief and discuss it in. Then ask each person to write down arguments, information, sources, claims, i.e. message tactics to change that Key Belief. Discuss everyone’s ideas in detail. When the discussion dies, ask everyone to write down any new ideas stimulated by the discussion. Repeat this process with each Key Belief.
2. Hire a Creative Team
I would recommend a dual track strategy with a Creative Team. Give them two sets of instructions. The first set would ask them to be creative and come up with what they think would be a good message tactic for changing the defined behavior. The second set would be to develop message tactics that specifically address the Key Beliefs (provide the examples generated by your focus groups – the creative team may laugh at this, but give it to them anyway). No matter what you do, a good creative team will give you several examples from Instruction Set 1. They will also be very enthusiastic about those creative executions. Be polite and consider them, then nonchalantly ask to see the executions for Instruction Set 2. If the executions from Set 2 don’t directly and clearly address the Key Beliefs, politely re-explain the concepts and ask the team to try again. When you have the executions from Set 1 and from Set 2, make sure you have paid the creative team, then hide the Set 1 executions (the groovy “creative” ones that may not address Key Beliefs) in a file drawer. Now, open Set 2 executions (that do focus on Key Beliefs) and look at them. As you evaluate them, ask one question: Does the message contain Strong Arguments that clearly, simply, and directly address the Key Beliefs?
While there is no magic bullet that automatically generates Great Messages, you can employ a very reliable empirical process to test and discover messages that stand a good chance of success. It is very simple and follows what a researcher would think of as an experimental (rather than survey) approach to data collection. You will systematically present message executions to a representative sample of people (Doers and NonDoers), have the respondents evaluate each message execution on key variables, and then select the best messages based on quantitative outcomes. No instincts, no gut feelings, no inspiration. Just perspiration, which Mr. Thomas Alva Edison thought was the royal road to invention.
1. Pick a small group of message executions.
It is not necessary that you have the final product in hand. Mockups in the form of storyboards, cartoons, PowerPoint slides, etc. are sufficient. The execution has to contain the key words and images and it should be in roughly the same format as it would appear in the final product (for example, a series of PowerPoint slides with voice and sound would simulate a TV ad much better than a written narrative description of a TV ad; another example, again a PowerPoint series could fake a website better than a cartoon storyboard depiction). Obviously, the closer you can get to the final format, the better the testing, but usually that final form will cost a lot of time, money, and effort and we’re just at the testing phase here.
2. Create an Audience Sample
A lot of people will run a few focus groups (total less than 30 people) on several message executions, then go with the executions that generate the most buzz. This provides yet another reason why most interventions have little or no impact. In this approach, one factor can wildly bias the response of a group (one charismatic group member, a careless group leader, a group misperception that they are supposed to find the “right” answer) and then mislead you into making a false judgment about the persuasiveness of a message. You need to find a relatively large sample of representative people, then have each individual view the message executions and provide reliable and valid ratings of each message. You need to have a larger sample (at least 30 people per execution) to ensure you’re getting a heterogeneous response. You need to run the testing at the individual level rather than the group level because overwhelmingly that how’s people get messages. It has much greater ecological validity and is much less susceptible to the contamination effects that easily occur in groups. You are much more interested in testing messages on the NonDoers, so it is smart to overweigh the sample (two NonDoers for everyone one Doer is fine). The strategy here is to find messages that move the NonDoers without negatively affecting the Doers. A smaller sample of Doers will reliably detect that large negative effect message while you need a larger sample of NonDoers to find a good message with moderate effects.
3. Test the Executions
Let’s say you’ve got 3 good candidates to test. You will need at least 90 total people to view them. You can get more, but once you start getting over 40 or 50 viewers per execution you are wasting your time and money. (If you understand power analysis and the practical impact of messages you realize that you are looking for messages that generate moderate to large effect sizes, thus, you don’t need huge samples. By contrast in the TpB testing, you are looking at so many variables and making breakouts that you need more sample size to generate enough power to test small effects.)
Randomize the order that you present the executions to each person. You then follow a simple action plan of: 1) present message, 2) get quantitative rating, 3) repeat until you’ve shown all executions.
4. Rating Items
The message provided new information to me.
Strongly Agree Agree Uncertain Disagree Strongly Disagree
1 2 3 4 5
The message was relevant and interesting to me.
Strongly Agree Agree Uncertain Disagree Strongly Disagree
1 2 3 4 5
I liked the message.
Strongly Agree Agree Uncertain Disagree Strongly Disagree
1 2 3 4 5
Then ask about the Key Beliefs. Imagine that your TpB research had determined that there were 2 Key Beliefs on expense and taste.
Fruits and vegetables cost too much for me.
Strongly Agree Agree Uncertain Disagree Strongly Disagree
1 2 3 4 5
I really do not like the taste of most fruits and vegetables.
Strongly Agree Agree Uncertain Disagree Strongly Disagree
1 2 3 4 5
5. Quantitative Analysis
You want to find messages that score the highest on all of the processing items. It is rare to find one message that scores best on all criteria, however. The simplest way to do this is run a one way discriminant analysis with message as the independent variable and then each process item as the dependent variables. You could also run a series of one way ANOVAs. You are simple looking for the message or messages that clearly produce high means (on the scale we’re using anything less than 3 is okay with scores getting closer to 1 being better). Clearly the most important test focuses on the Key Beliefs. If people are reporting much more favorable beliefs after viewing a particular message, then even if the message is not new or not particularly likable or interesting, don’t get too worried. It hit the Key Belief.
6. What If Nothing Works
The silence is deafening. Nothing works. Now what?
If you have the time and money, go back to the message creation step, and start over. If you don’t have the time and money, take the feedback on all the messages and make a stew, that is, combine what appears to be the best elements from some messages and hope for the best. From my experience in doing this, talking with others who do it, and reading a lot of research, failures typically occur because you got too cute, clever, or inspired. In other words you got emotional. If you keep a strong rational focus, you should be okay. That means do the TpB survey by the book, find those Key Beliefs, then construct messages that clearly, simply, and directly address the Key Belief. Please review the text we used in the “1% or Less” and “Wheeling Walks” campaigns. Hardly Shakespeare. Nothing that would win an award. But realize that each campaign created the largest reported behavior change in its domain ever published.
Once again, if you are working alone (class assignment, underfunded work project), don’t get discouraged with all the details here. Get the concept, then cut it down to your needs and resources. You can’t do all this testing unless you are a wild eyed locust eating maniac from the desert who’s got to know the Truth. Okay? Focus, here. Read all these details and figure out how you can apply them to your circumstance. The point is to stay systematic and theory based.
You should now have a couple of messages for each behavior definition that have good empirical support. You’ve got evidence that “this” message changes Key Beliefs and is well liked while “that” message lacks good evidence. This should be simple. Pick the messages that have empirical evidence for producing change. There will be other messages that you really, really, really like and find compelling. That’s nice. Put them aside and move on with your science.
Action Step 4: Generate An 80% Reception Plan
Goal: Create Message Distribution Plan That Reaches At Least 80% Of Your Target Audience.
A reception plan is not aimed solely at generating awareness. A reception plan is aimed at making sure a large majority of your audience engages in many instances of high WATT processing about strong arguments directly addressing key beliefs. An intervention that gets a lot of reception (or exposure, reach, awareness, recognition, etc., in other words did you get the message?) but not much behavior change clearly fails. This is not a tricky, subtle, or esoteric point. You can easily deceive yourself into thinking a communication intervention is working because it generates awareness. Actually all you are doing is fooling yourself with semantics. When you make a media buy for a campaign you are literally buying awareness, exposure, reception, etc. Thus, when a survey comes back with 80% of respondents saying they saw your message and you had made a “large” media purchase you have achieved nothing unexpected. You described what you wanted (80% reception) and the media buyer used well established and proven formulas for hitting that mark. It’s like having a chicken dinner and then going through the garbage to determine what people were eating and discovering a lot of chicken bones.
Most interventions rely upon interpersonal contact rather than mediated contact to distribute intervention messages. If you are doing a truly “public” intervention where you are trying to influence a population rather than a segment of a population, interpersonal approaches are almost required to fail because it costs so much to produce enough workers to reach all the population and many workers are not particularly skilled as communicators. By contrast, mediated distributions are cheaper by far and can be controlled by people with the required communication skill.
To obtain a good reception plan the best approach is to hire a media company with a proven track record. There is a surprising amount of science behind reception plans. Media companies have been measuring the relationship between the media action (how many messages appearing in how many channel and vehicles) and outcome (sales of products and services, Nielsen rankings, opinion surveys, etc.) for decades now. They can reliably tell a campaign planner with a high level of certainty exactly how many times a message needs to run to achieve various levels of reach. They can also reliably tell you how to target messages to specific population segments by picking specific media channels and vehicles. A good planner will provide an explicit schedule that details when, where, and how often a particular message will be distributed and will then provide an after action report that details a estimate of how many people got the message. When selecting a media company, look for these kind of details in their track record.
You may also want to share your campaign timeline with the media planner and discuss creative possibilities with your buying plan. For example, it might be more effective to begin each new target behavior with a massive media buy aimed at creating a general level of awareness and thinking in the population. Then, run very tailored messages on specific channels and vehicles to reach specific segments using a smaller buying plan. A media planner needs to get a sense of the rhythm and variation in your timeline to make a smarter buy.
You might recall an old quote about money attributed to the late Senator Everett Dirkson who said, “A billion here and a billion there and pretty soon you’re talking real money.” His point was that the sheer size of a budget sometimes makes people careless in their planning and use of money. This point can be kept in mind when dealing with media budgets. To get into this game you have to have a lot of money compared to the typical intervention. You need to keep your focus with this kind of quantity because it is easy to let quality slip. First, a media dollar should buy reach (that 80% awareness criterion). Second, a media dollar should buy processing (opportunities to generate favorable thoughts about the arguments).
It is fairly easy to quantify when your media buy is hitting the 80% criterion. I do not think that spending more money to get past the 80% value is worthwhile because if those 20% haven’t gotten the word, they probably do not want to hear about it at all. This 20% is composed of the segment of the population that does not have a TV, radio, newspaper, magazine, or computer and does not hang out in places that lack these media. You could more profitably spend those potential media dollars on interpersonal contacts with these isolated or resistant people.
It is hard to quantify when your media buy has generated sufficient opportunities for most people to have carefully and effortfully thought about your arguments. My experience indicates that the 80% reception criterion correlates with sufficient processing opportunity. Some research indicates that when people have gotten a message 3 times, they have probably engaged in all the high WATT processing they can give to this topic. Repetition won’t help.
If you are doing a smaller project (the class assignment or underfunded work project), you will probably not be doing the major media buy outlined here. But you still face the same problem. You want to make sure that your targets hit that 80% reception mark. And you know that you need to use Placement, Frequency, and Contrast to accomplish this. As always, don’t get bogged down in the details here if you are doing a smaller project. Just understand the principle and cut it down to your needs and resources. Chances are good that you’ll be doing mainly interpersonal communication (face to face), so you’ll be in a much better position to understand the WATTage of your targets as you make sure they get the arguments. You’ll see how the targets react to the information and will be able to adjust new messages as needed. Usually, in smaller projects, the Reception part of the intervention is a lot easier than with large projects. As always, stay systematic and theory-based.
The reception plan accomplishes two highly dependent goals. First, it generates a lot of awareness. Second, it gives your population many chances to engage in high WATT processing of strong arguments aimed at key beliefs. The way you achieve these goals appears to be simple: Get a lot of money. While it is true that communication campaigns can cost a lot of money, it is crucial that you think carefully about your media dollars. You should get the biggest bang for you buck. And, again, the biggest bang is a reception plan that get both awareness and opportunities for high WATT processing.
You now have a detailed blueprint for rolling out a communication intervention that has a reasonably good chance of practical success. You have defined your target behaviors, identified the key beliefs that drive these behaviors, developed strong arguments that change those key beliefs, and created a reception plan to reach over 80% of your population. You have used state of the art theory and research throughout the development process and you can now use the same theory and research methods to assess the intervention. Think about how easy it will be to collect your data, analyze it, and then write it up. You’ve got a great template that guides creation, implementation, assessment, and write up. Nothing is wasted or isolated or weird. It all fits together in a coherent, scientific, and practical whole. It’s almost paradigmatic.