Message discipline is a common characteristic of winning candidates, especially in races that fall below the top of the ticket. A voter chooses a candidate because that candidate gives him or her a reason to do so. But what good is message discipline if the candidate is communicating weak messages or is communicating to the wrong voters? Testing message strength and targeting are, therefore, important.
The traditional approach to message testing has simply been to ask survey respondents a block of push questions then compare the topline scores across the block. Additional analysis consists of crosstabulating individual messages with demographic variables. This article challenges that tradition, arguing that Interactive Analytics facilitates a superior theory of message strength. Six properties of messages explain why:
1. Message effectiveness is defined by voter migration
A message is effective only if it causes voters to migrate from a previously held position to a more desirable position. In a survey this migration might be measured by asking respondents a first ballot question, testing a block of messages then asking a subsequent ballot question.
The traditional model of message testing is limited to examining how the survey sample responds to the block of messages. But we learn little from those who consistently vote the undesirable (or desirable) position on both ballot tests. Rather, our real interest lies with those who migrate in one direction or the other. Interactive analytics lets us isolate and analyze those who migrated as a result of exposure to messages.
2. Messages possess multiple dimensions
Message effectiveness might also require examining multiple message dimensions. Consider a candidate’s stance on an issue. A voter might agree with the candidate but dismiss his or her agreement as less than motivating.
To impel voting behavior the voter must believe that the issue is important, the candidate’s claim is believable, and that the candidate’s argument withstands an opponent’s rebuttal. A failure of any one of these three tests leaves the voter unmoved. The task of message testing then becomes one of measuring the dimensions of importance, believability, and resistance to rebuttal.
Find a message that passes all three tests and you’ve got an effective message. Interactive analytics visually overlays parallel blocks of messages, with each block measuring a different dimension. This makes multi-dimensional analysis straightforward.
3. Campaigns cast messages at different levels of abstraction
We often speak as if messages exist on a single level of abstraction. They don’t. At the highest level of abstraction messages consist largely of labels: conservative, taxpayer advocate, issues candidate, etc. Moving down the ladder of abstraction brings us to specific arguments that give voters a reason to vote for a candidate; below that are specifics such as narratives, evidence, and facts.
Testing candidate positioning involves testing at that highest level of abstraction. Interactive analytics allows for candidate positioning tests using paired-comparison methodology. Each candidate position is contrasted with every other position and each position is scored. Those scores can then be compared across levels of any variable measured in the survey.
4. The other side in the campaign has messages too
We test not just our own candidate’s potential messages but the opponent’s messages as well. So how do we compare message strength across campaigns?
Interactive analytics calculates index values. Question responses can be assigned as positive, negative, or neutral. Index values are then calculated by subtracting the percentage negative from the percentage positive then adding a constant of 100. Interactive analytics uses index values for a number of purposes; one is to let users to visually and quantitatively compare message strength across campaigns.
5. Message effectiveness varies across segments of the voting population
Messages are typically not universally effective; hence the necessity of selectively targeting messages. Interactive analytics lets users sequence through selected target groups while examining message blocks.
6. Messages must be considered in anticipation of changing campaign scenarios
We know that message strengths are not static, but dynamic. How, for example, would messaging change if voter turnout drops below expectation or voter turnout drops for one party but not the other? The possible scenarios are endless. Interactive analytics allows for what-if simulations of different scenarios while performing any of the analytic functions.
As should be obvious, message testing using toplines and crosstabs is archaic. But as James A. Michener put it, “An age is called Dark, not because the light fails to shine, but because people refuse to see it.” If you would like to take a look just give the author a phone call.
Val Smith, Ph.D. has been a political pollster for over 35 years, and is a Principal at Data Analysis and Display, LLC, authors of Porpoise Survey Analytics and Orca Data Editor. Val can be reached at (916) 932-2374 or valsmith@PorpoiseAnalytics.com.