Demographic questions should be fairly easy to ask as everyone knows the right answer for themselves. But it’s not always so simple. Here are a few tips for asking those ‘simple’ questions that will help you elicit the most valid answers from respondents.
Gender: This is perhaps one of the easiest questions to ask. At least, historically it has been one of the easiest questions to ask.
Nowadays, society and our culture has come to understand that there are more nuances than simply ‘male’ and ‘female,’ and regardless of how you feel about those nuances, you need to provide answer options that will encourage your respondents to answer truthfully.
With that in mind, think about whether you must use the word gender or sex. In many cases of market and social research, simply avoiding the use of both terms will avoid the associated confusion of interpreting those terms. For instance, it is probably essential to ask about sex when conducting a study of aging and hormone treatments, but must you know the gender or biological sex of a person who buys milk or toothbrushes? Tweet this!
Also, by providing additional answer options, we can accomodate people who genuinely are unsure about how to classify themselves (remember, biology isn’t always clear cut!). It’s better to know which data are worrisome as opposed to letting invalid data hide within valid data!
(By the way, this technique also works well for questions about race/ethnicity. If you aren’t sure whether to ask about race or ethncity, consider whether you must use one of those words at all.)
Example: Which option describes you? / Which answer describes you?
Female
Male
Other
Prefer not to answer
Age: There is no ‘correct’ way to ask people their age. There is only a respectful way. Consider whether you truly must know their exact age. Must you ask people to type in their exact year of birth? Must you present people with a long drop-down list of ages or years? These options might give you as the researcher more flexibility with your analyses but, at the same time, they increase the cognitive load for every person, they increase anxiety levels for some people, and they increase the difficulty of actually completing the task particularly on small mobile devices. At the end of the day, respondent experience and data validity must come ahead of the researcher’s best case data analysis scenarios.
Where possible, offer grouped age options. And when building these options, be sure to consider ages that are important culturally. Consider whether legally defined ages of children, legal voting and drinking ages, or retirement ages might be related to your research topic and incorporate those ages as needed. Tweet this!
Example: Which age group describes you?
14 or under
15 to 17
18 to 29
30 to 39
40 to 49
50 to 59
60 to 69
70 to 79
80 or over
Employment: I’m going to assume that you went to college or university. Did you have a part-time job at the same time? Were you trying to start a business at the same time? Were you responsible for a baby or toddler during that time? Were you caring for an aging parent or grandparent at the same time? What i’m getting at is that it’s time to stop asking the employment question as if people who are employed do not have families, do not hold second jobs, and do not attend school. And similarly, as if students are not starting businesses and caring for elderly parents or relatives.
This problem is easily solved by ensuring your employment question is a ‘select-all’ format rather than a ‘select-one’ format. When it comes to data analysis, simply filter the type of people you wish to understand, whether that be people who are employed full-time or people who are caregivers, and run the analysis. In fact, using this format will give you additional analytical options. I’m puzzled why people haven’t been doing this all along!
(By the way, this technique also works well for questions about race/ethnicity. People reflect long historical and cultural paths that simply aren’t homogenous.) Tweet this!
Example: Which of these describe you?
Full-time employed
Part-time employed
Not employed for pay
Caregiver (e.g., children, elderly)
Homemaker
Full-time student
Part-time student
Other
Income: This is a tricky question to ask because some people feel as though their income is an expression of their personal worth. They might be embarrassed to share their income, even though our questionnaires are anonymous, for fear that someone might find out their income.
Because of this, it is extremely important to include a ‘Prefer not to answer’ option. As with other demographic questions, it is vastly more important to collect valid data than precise data. When we can separate aspirational or exaggerated responses by providing a ‘Prefer not to answer’ option, our data and subsequent analyses will better off.
Further, when choosing break points for income questions, think about the natural distribution of incomes. Most people earn less than $100 000 and a very small percentage earns more, so provide more breaks in the lower ranges and few in the higher ranges. Also consider your target group and the users of the category. If your target group is likely to have higher incomes, then adjust the breaks to reflect that. Tweet this!
Example: Which of these describes your income last year?
$0
$1 to $9 999
$10 000 to $24 999
$25 000 to 49 999
$50 000 to 74 999
$75 000 to 99 999
$100 000 to 149 999
$150 000 and greater
Prefer not to answer
We’d love to help you design great quality demographic questions. Please get in touch with us!
You might like to read these:
- When Smell-O-Vision lets you down, a sensory test will cheer you up!
- Stories from the Research Field Director: It’s raining turkey jokes!
- Communication isn’t a carefully crafted questionnaire
Sasha Haig is an Account Manager at Canadian Viewpoint. She loves tackling the complex multi-modal studies while working collaboratively with clients to ensure their research objectives are met, and looking for new ways to approach data collection. In her free time, she likes hiking and sharing her unique tastes in cooking on her food blog.
Gender: This is perhaps one of the easiest questions to ask. At least, historically it has been one of the easiest questions to ask.
Nowadays, society and our culture has come to understand that there are more nuances than simply ‘male’ and ‘female,’ and regardless of how you feel about those nuances, you need to provide answer options that will encourage your respondents to answer truthfully.
With that in mind, think about whether you must use the word gender or sex. In many cases of market and social research, simply avoiding the use of both terms will avoid the associated confusion of interpreting those terms. For instance, it is probably essential to ask about sex when conducting a study of aging and hormone treatments, but must you know the gender or biological sex of a person who buys milk or toothbrushes?
Also, by providing additional answer options, we can accomodate people who genuinely are unsure about how to classify themselves (remember, biology isn’t always clear cut!). It’s better to know which data are worrisome as opposed to letting invalid data hide within valid data!
(By the way, this technique also works well for questions about race/ethnicity. If you aren’t sure whether to ask about race or ethnicity, consider whether you must use one of those words at all.)
Example: Which option describes you? / Which answer describes you?
Female
Male
Other
Prefer not to answer
Age: There is no ‘correct’ way to ask people their age. There is only a respectful way. Consider whether you truly must know their exact age. Must you ask people to type in their exact year of birth? Must you present people with a long drop-down list of ages or years? These options might give you as the researcher more flexibility with your analyses but, at the same time, they increase the cognitive load for every person, they increase anxiety levels for some people, and they increase the difficulty of actually completing the task particularly on small mobile devices. At the end of the day, respondent experience and data validity must come ahead of the researcher’s best case data analysis scenarios.
Where possible, offer grouped age options. And when building these options, be sure to consider ages that are important culturally. Consider whether legally defined ages of children, legal voting and drinking ages, or retirement ages might be related to your research topic and incorporate those ages as needed.
Example: Which age group describes you?
14 or under
15 to 17
18 to 29
30 to 39
40 to 49
50 to 59
60 to 69
70 to 79
80 or over
Employment: I’m going to assume that you went to college or university. Did you have a part-time job at the same time? Were you trying to start a business at the same time? Were you responsible for a baby or toddler during that time? Were you caring for an aging parent or grandparent at the same time? What i’m getting at is that it’s time to stop asking the employment question, as if people who are employed do not have families, do not hold second jobs, and do not attend school. And similarly, as if students are not starting businesses and caring for elderly parents or relatives.
This problem is easily solved by ensuring your employment question is a ‘select-all’ format rather than a ‘select-one’ format. When it comes to data analysis, simply filter the type of people you wish to understand, whether that be people who are employed full-time or people who are caregivers, and run the analysis. In fact, using this format will give you additional analytical options. I’m puzzled why people haven’t been doing this all along!
(By the way, this technique also works well for questions about race/ethnicity. People reflect long historical and cultural paths that simply aren’t homogenous.)
Example: Which of these describe you?
Full-time employed
Part-time employed
Not employed for pay
Caregiver (e.g., children, elderly)
Homemaker
Full-time student
Part-time student
Other
Income: This is a tricky question to ask because some people feel as though their income is an expression of their personal worth. They might be embarrassed to share their income, even though our questionnaires are anonymous, for fear that someone might find out their income.
Because of this, it is extremely important to include a ‘Prefer not to answer’ option. As with other demographic questions, it is vastly more important to collect valid data than precise data. When we can separate aspirational or exaggerated responses by providing a ‘Prefer not to answer’ option, our data and subsequent analyses will better off.
Further, when choosing break points for income questions, think about the natural distribution of incomes. Most people earn less than $100 000 and a very small percentage earns more, so provide more breaks in the lower ranges and few in the higher ranges. Also consider your target group and the users of the category. If your target group is likely to have higher incomes, then adjust the breaks to reflect that.
Example: Which of these describes your income last year?
$0
$1 to $9 999
$10 000 to $24 999
$25 000 to 49 999
$50 000 to 74 999
$75 000 to 99 999
$100 000 to 149 999
$150 000 and greater
Prefer not to answer
We’d love to help you design great-quality demographic questions. Please get in touch with us!
