How to create a survey without the respondent being affected by slanted words or images
When you have got a result, you do not want to start questioning the method that has been used but instead want to be able to rely on it being well-founded and produced in a reliable way. It is therefore important that the researcher has not made use of words or images that are loaded or slanted and thus have impacted on the results in one way or another.
An image that affects the whole result of the survey
An example which tested the impact of pictures on outcome was provided by a survey about perceptions of health. The statement “I believe that I have good health” was tested on two different sample groups. The two groups were each given images with the question in the survey: One of them portrayed a sick woman in bed, while the other showed a woman training at a gym. Perhaps unsurprisingly, the different groups gave different answers. The group who received the image of the sick woman felt significantly better than the comparison group. The reason? In both cases, the respondents compared themselves to the image and estimated their health higher based on a bed-ridden person and lower based on a person training at a gym.
Avoid unintended influence in surveys
[bq_right]Do not influence the respondent by the use of charged words or images[/bq_right]
Images can be a good complement in many surveys, but in order to avoid any unintended impact on the result it is important to use either neutral images, or both types of image to avoid pushing the respondent in one direction or the other. The same applies in the case of word choice. If you invoke a sense that it feels good or less good to answer a question in a certain way, the results will lose credibility.
For example, the wording “Zlatan believes that Barcelona play the most beautiful football in the world. What do you think?” is clearly slanted. For those who are not conversant with different teams’ ways of playing football, it is easy to agree with someone with greater knowledge in the area. If, in the example, there is no trust in Zlatan as a person, there will be many who disagree with his statement simply because it was voiced by him. Therefore, you should avoid all loaded lines of reasoning related to a question. You then give the respondent the chance to communicate their own opinion as they understand it.
What do you think?
Have you experienced difficulties in remaining completely neutral when it comes to wording and images in surveys? How much influence do you think it has on the result? Feel free to discuss this with me in the comments section below, or email me at firstname.lastname@example.org .