An increasing large proportion of Americans are climbing into the “over 65 years of age” bracket, which means that this demographic is becoming ever more interesting to researchers of all kinds.
For nearly every discipline, this means they need to survey older adults regularly.
Ph.D. candidate Kelly Quinn of the University of Illinois at Chicago points out that, “Surveys are a common tool to gather epidemiological, behavioural and social data on all populations, including older persons. And like other types of research, it may be prone to specific weaknesses depending on the processes used to select samples, gather data and interpret results.”
As survey creators, we can help mitigate these concerns by giving special consideration to survey design, administration, and data analysis whenever older adults may be among our respondents.
Survey Design for Older Respondents
The Pew Research Center estimates that around “two in five seniors indicate that they have a ‘physical or health condition that makes reading difficult or challenging’ or a ‘disability, handicap, or chronic disease that prevents them from fully participating in many common daily activities’.”
This means that we need to create survey components that allow older adults to interact with our surveys easily and without frustration.
Keep these tips in mind even when your target audience doesn’t include older people; they make your survey better even for respondents of all kinds:
- Use high contrast colors throughout your survey.
- Make sure to use large fonts for question text, answer options, instructional text, and buttons.
- Button sizes should be larger than normal, with plenty of padding between clickable areas.
High Contrast Colors in Survey Design
Light colors may produce visually pleasing combinations, but on small screens or for respondents with deteriorating eyesight, they can be a serious impairment.
Stick with saturated colors, and be sure your text and your background are very distinct.
When in doubt, black text on a white background is the way to go.
Using Large Font Sizes
The smallest font size you should use for any text that might be read on a mobile device is 16 pixels.
If you know you’ll be surveying older adults, you may want to increase that minimum to 18 or 20 pixels so that respondents don’t have to squint or zoom in.
You should also ensure that there is enough space between the lines of text to allow respondents to read it easily. Typecast recommends 1.25em for phones and small tablets, and 1.375em for large tablets.
Obvious Buttons for Survey Actions
Older respondents simply don’t have the decades of computer interactions that younger generations are accustomed to. This means you’ll need to make sure any in-survey interactions, such as a “Submit” or “Next Page” buttons, are very clearly marked.
The simplest way to do this is to follow the high contrast rule again: make sure your button color is markedly different from both the text overlay and the background.
To avoid frustration on mobile devices, your button should be at least 48 pixels wide. If you have multiple clickable areas near one another, they should be 48 pixels away from each other to avoid accidental clicks.
Considerations for Survey Content
The content of your survey can have just as much of an impact on its accessibility as its design.
In Cognition, Aging and Self-Reports, Fergus I.M. Craik states that elderly respondents:
“may have particular difficulties in comprehending questions, especially, perhaps, in telephone interviews; they may be more liable to impose their own assumptions on what is being asked and why; they may have difficulty in holding and integrating the successive parts of long or complex questions and they may have problems in remembering details of their own habits and actions.”
When designing survey questions, it can help to ask about what he calls, “large-scale salient events,” rather than smaller details.
For example, “Where did you take your vacation last summer” rather than, “What was the name of the hotel?”
Reaching Older Adults With a Survey
Once you’ve carefully constructed your survey project, it’s time to get it into the hands of respondents. Online surveys can still reach older adults, but they may not provide precisely the same data as the same surveys taken by younger respondents.
The Pew Research Center found for the first time in 2012 that more than half of older adults (defined as those 65 or older) were internet users. As of 2014:
59% of seniors report they go online — a six percentage point increase in the course of a year — and 47% say they have a high-speed broadbrand connection at home. In addition, 77% of older adults have a cell phone, up from 69% in April 2012.
Although more and more older adults are going online, that doesn’t mean that their attitudes about taking online surveys are identical to those of younger respondents. Americans over the age of 65 tend to have a marked reluctance for providing confidential information over the internet, as well as a very different understanding of the value and function of technology in a cultural sense.
Quinn points out that as, “researchers introduce more technologically sophisticated tools into their research habits, understanding of cultural distinctions becomes increasingly important because of their potential significance to the quality of survey results.”
Distribution method in particular need to be carefully considered to get the best possible response rate and reduce response bias. Offering surveys online is the simplest approach, but many studies have found that a variety of options for providing data is the best approach.
This may include adding face-to-face surveys, paper surveys, or telephone surveys to your project.
Response Rates in Mailed Versus Hand-Delivered Surveys
Despite increases adoption in technology use, there remain special considerations when surveying older adults.
For example, a study on survey methods and response rates among rural community dwelling older adults found a significant improvement in response rates when surveys were delivered by hand rather than in the mail:
The overall survey response rate was 43.9%. Older adults who were handed the survey packets by the home-delivered meals drivers were older and significantly more likely to return the survey (57%) compared those who received survey materials in the mail (31%).
You may not want to hand deliver all of your surveys, but consider the demographics of your respondents carefully when choosing a distribution strategy.
Possible Bias in Older Respondents
Once you’ve designed an accessible survey and carefully distributed it to your respondents, you still need to consider how your respondents’ ages may introduce bias into your data.
In her paper, “Methodological Considerations in Surveys of Older Adults,” Kelly Quinn reminds us that, “As survey methodologies grow more technologically-dependent these variances become a potential source of bias and an increasingly important component of survey data analysis.”
Even if you reach a large number of older adults, your data may not be representative of that population on a larger scale. Quinn continues:
“survey researchers who use technologically sophisticated data collection measures must consider whether their respondent base of older adults adequately represents the characteristics of older adults in the broader population: older adults sufficiently comfortable with survey administration technology may not be representative of older adults generally.”
Two Options for Reducing Bias in Survey Data
For those creating surveys aimed at the older adult population, there are two simple ways you can reduce the response bias that can contaminate survey data.
- Capture health status up front, just as you would basic demographic data such as age.
- Be clear about how you will protect the privacy of individual responses.
Health Information as a Demographic Question
In her paper, “Methodological issues in aging research: An introduction,” KW Schaie suggests that we ask about health status in the early part of a survey, even if it’s not a research variable under consideration.
This is particularly important if you’re conducting any type of longitudinal research that will follow respondents over time.
Asking this type of question up front, “allows researchers to control for the incidence of illness and physiological disability in the data analysis.”
Communicate About Privacy
Online privacy is a concern for nearly anyone who provides confidential information in a survey, but for older adults privacy worries can drive them to avoid responding to surveys at all.
Do your best to assuage these fears by including clear information about how you plan to safeguard responses in your survey instructions.
Keep in mind that older adults often assign different cultural importance to online interactions than their younger counterparts. They most likely won’t be inclined to offer up answers simply for the sake of online data collection. Be as clear as you can about your plans for the survey data and what steps will be taken to anonymize answers (if appropriate).
Survey Older Adults With Confidence
Gathering data from the aging population will continue to be an important component of research in almost any field.
While it can come with its own unique challenges, by approaching survey design, distribution, and data analysis carefully we can still draw accurate conclusions from our results.