In the last module, we talked about how do you highlight the benefits of participating in a survey, to encourage people to respond to your request for surveys. In this module, what we're going to talk about is how do you reduce the cost of participation in the survey? So we're going to look at some specific strategies for increasing response rates to your survey by making it easy to use. Remember we talked about social exchange, that when we're asking people to do any volunteer activity, including participate in a survey, they'll consider their effort in terms of the cost/benefit ratio, right? Is cost of participating this activity worth the benefit that I get from it? We talked about lots of ways you could increase benefits, but how do you reduce the cost of participating in a survey? We think about cost in terms of effort, right? How effortfull is it for your potential participant or respondent to participate in the survey that you're asking them to do? In that way, we're really lucky a lot of stake in this module are UX researchers. And designing a survey is like most other user experience tasks, we just need to be user focused. We need to think about designing a survey like we think about designing any other interaction that we want a person to do. So, the primary way I think that a lot of researchers lately have tried to reduce the cost or effort of participating in a survey, is by reducing the burden of length, right? And that's mostly by making the survey much shorter than we would have seen in the past. Especially with digital communication, people are much more accustomed to short interactions and very quick interactions. Having a survey questionnaire that's only one item would have been unthinkable decades ago, but is now very common in certain types of online research. You want to avoid, even if you're using more than one question, kitchen sink questionnaire designs. This is another rookie mistake I've seen for survey designers, where they're not quite sure what they want to ask. So you just throw in every question they can think of to make sure that at the end of the day, they're going to get something useful out of their survey. That's not the way to do it in terms of reducing the length of the survey for your respondents. Now, while those one-question questionnaires have become more popular and you know why? Because they basically make it very easy for a person to participate. There is a tension between the length of a survey and quality of the data you get from questions. A one-question survey has a lot of benefits in terms of increasing response rates. However, it's going to be limited, in terms of the amount of data that it provides for insights later down the road. In the section that we have on asking questions, we're going to be talking quite a bit about why would you ask multiple questions in the first place? What are the tradeoffs between asking a single item versus multiple items? And one tradeoff, of course, is just are people going to be scared away by answering multiple questions. This is kind of the bargain that we have to make sometimes, especially when we're in a mode like online surveys where it's very hard to get people to comply with a request to do a survey. Another way we can reduce the effort of participating in a survey is to reduce the complexity of the survey that we have. Now complexity can come from multiple different sources. It can include things like big skip patterns or complicated designs of your survey. Some very high end surveys often include things like data collection from outside sources, pulling records from your medical file or things like that. In general we're not going to experience a lot of those. Where I find a lot of complexity in the kind of more day to day survey world is through using very specific or hard to answer questions, right? If I'm asking about very specific behavior or very specific financial information or employment information. If I ask for instance, list all of the places you've lived for the past 15 years. That can be data that's very hard to sum it up, and can be very disincentivizing for the respondent because it makes a lot of complexly for that in terms of answering the question. A common area where I see complexity added for the respondent is what I think of as another rookie mistake, which is including an open-ended question in your survey after every closed-ended question. An open-ended question it would be something like, why do you think that? With just a text box where a person is answering that. Now I could see why our surveyor would want to do that, you're trying to get more at these why questions and you're trying to cover, in case you didn't ask the right question. But every open end adds considerable burden to your respondent. If they see that kind of complexity, it's going to really disincentivize a person from participating in the first place. This may seem obvious, especially to a group of user experience researchers, but it's very important to use good visual design principles when thinking about a survey. Now of course, there's the aesthetics of that. Clear readable fonts. You want to think about color and layout choices. In a later module, we're going to talk about different survey tools that you can use and how they make this design easier or harder. But you also want to think of your ease of use principles when you're thinking about designing visually a survey. How many pages should you have in your survey? Is it one page or is it multiple pages? How much does a person have to scroll down a page to get to all the questions? Is there a breadcrumb trail, so they know how long they're in the survey? Are there lots of matrix questions? Matrix questions are where you have basically the same scale and then lots of different questions that use that same scale. Those can be really confusing and hard to answer. How you design a survey can play a huge role in how a person perceives their desire to participate in it. This one is a little less obvious than do good design, but you want to avoid subordinating language in invitations. By subordinating language what we mean is, you want to avoid an invitation where it sounds like a person has to respond or that there is not the social exchange that we're trying to incur here. So a good example of a subordinate invitation is, for us to help improve the product, it is necessary for you to complete this questionnaire. We will flip that route and add something that wasn't in that subordinate frame. We would say, will you please be part of helping us to improve this product? You can see the difference, right? Instead of making it the respondents responsibility, now what we're doing is we're asking them for the help and we started the reciprocal relationship. This also might seem obvious, but you want to make it easy to respond. Some modes, like mail or phone, it can be harder to send data in than others, like web or email. The easier it is for a person to send in the data to you, the more likely it is are going to comply with your request. Is one thing to they just have to go to an online space and fill out a survey online. It has considerable burden if you are mailing something to them, they have to fill out the survey which has its own costs and then find an envelope and mail it back to you. The more cost there is to actually responding, the less likely people are going to respond to you When you're picking a mode, you want to make sure your matching that mode to the audience or the population that you're interested in. How do you choose between online, mail, or phone modes for instance? What are the preferences of your target sample? A lot of younger people for instance don't use phones that often for actual voice communication. So would a phone survey be the best way to get in touch with a younger audience? A lot of older people are less comfortable with texting. So would a SMS survey be the right choice for them? You want to think about the mode that you're using and match that to a prefered mode that your sample communicates with. We also want to minimize your request for personal or sensitive information. We talked about in a previous module, about how people will not respond to surveys if they, for instance, there's questions about financial interests or maybe it's health, their personal health or maybe it's about their children. Questions where people may be triggered to worry about their safety or the sensitivity of their data, can really make people not want to participate in your survey. So sometimes it isn't important to ask these questions and you're going to have to ask them, but you just want to make sure that you're not asking sensitive questions just for the sake of it. You want to make sure that you have tight research questions and good outcomes for your survey, so that you avoid asking questions that are just going to weird out your potential responders. So that's a good idea to provide an accurate estimate of the effort you're requesting. It's a good idea to say, is this survey going to just take a few seconds or just a few minutes? If it takes some longer time, you going to want to use a different mix of all the techniques we have been talking about. It is okay, for instance, if a survey takes a long time. It is just then maybe you got think about how we offer incentives in a different way or tokens of appreciation than if it is a very short survey. So the length of your survey and the actual honest assessment for how long that survey is going to take, should play a role in how you reassemble some of these techniques to increase responses. So in summary, one of the important ways to increase response rates in surveys is by reducing the costs of participation in those surveys. This folds in nicely when we thought about how to increase the benefits of participating in a survey. There are lots of different ways to reduce cost. A major way, for instance, is to keep the survey as short as possible to reduce respondent burden, but there are other ways. So for instance, making sure that your design is really strong. Making sure that you're not asking overly sensitive questions. Making sure that you're being respectful of people's time. It's important for us to think of designing a survey as a user-centered design activity. If we get ourselves into our respondents' mindsets, if we think about this again as a contextual exercise, that'll really help us to think about messages and about requests for participation that are matched to the needs of potential samples.
