In December, along with my contextual interviews, I tested out some prototypes. I use “prototype” here loosely, meaning that I showed participants images and texts related to bottled water, and asked them a series of questions about their reactions. Images and texts were judged separately by the participants, according to four factors (see note at end):
- Whether participants found the image/text interesting or uninteresting
- Whether the message was understandable or ambiguous
- Whether the participant felt encouraged to take action
- Whether the participant felt likely to change their attitude towards bottled water
Participants’ preferred images varied quite a bit. The images that stood out most, however, were two which showed plastic pollution on a beach and in the ocean, and one with a message about the amount of oil used to produce plastic bottles. These images are somewhat polar opposites: the first two are highly emotional; the latter is based on logic and statistics. The images of people drinking out of or holding reusable bottles sparked some interest, but were overall judged to be too ambiguous, and unlikely to lead to action or an attitude change.
As a side note, I realize now that including the images with text did not allow me to focus only on the visual quality of the images. If I were to do this activity again, I would separate out images which also contained text.
Download Text A (longer text with statistics)
Download Text B (longer text, first-person)
After viewing the images, participants read 4 short texts and 2 longer texts. In general, they reported liking the shorter texts, but being more convinced by the longer texts. This is a dilemma that I believe can be solved by giving audiences a small amount of information and allowing them to choose more, if desired.
Participants unanimously preferred the texts with statistics (such as Text A), and disliked Text B, which was written in first-person and had no statistics. Some people said that Text B felt like “one person’s decision” and reacted by saying “so what?” One person said that it was too overtly persuasive; a few others called it “preachy.”
The results of these prototype evaluations have led me to hypothesize about possible design recommendations. My recommendations can be divided into two categories: qualities that the communication piece must contain, and qualities that the communication may contain.
Possible design recommendations
Communication piece must contain:
- Carefully-chosen statistics given in context. Participants unanimously preferred messages with statistics, over similar messages without statistics.
- A message which does not feel “preachy.” The first-person text turned many participants off. This quality must be evaluated on a case-by-case basis.
- A concise message. Information must be easy to digest and remember. Longer texts should be broken up into chunks.
- Specific actions to take. Information about a problem must be accompanied by concrete solutions which the audience knows they can do (see New Tools for Environmental Protection)
- Assurance that individual actions will make a difference. Some participants expressed doubt that their actions would lead to change.
Communication piece may contain:
- An emotional tone. Some participants were most drawn to images and texts with an emotional appeal.
- Practical advice. Other participants were most drawn to images and texts that showed efficiency, or cost or energy savings.
- New information for the viewer. Some participants were most interested in facts of which they were unaware, such as the amount of oil used in plastic bottle production.
- The option to learn more. Participants liked short texts but were more convinced by longer texts. Audiences might be given an option to dig deeper into the issue.
- A role model performing the desired behavior. Increasing the personal relevance of a message motivates message scrutiny (see The Psychology of Attitudes and Attitude Change). But participants expressed strong feelings of distaste when shown images of people with whom they couldn’t identify; similarly, the anonymous first-person narrative in Text B triggered negative feedback.
During the rest of January and February, I will be creating and testing a more robust prototype in order to test my hypotheses. I hope this will lead to more clearly defined design recommendations.
Note: the questions about the prototypes were taken from William James McGuire’s “Communication-Persuasion Model as Input-Output Matrix” (as shown in Public Communication Campaigns, 1989). McGuire’s matrix is a classic theory, and breaks down the steps that a persuasive message must go through in order to lead to behavior change.