How to motivate respondents to take surveys? - HuginX

How to motivate respondents to take surveys?

We know very little about why people participate in research and even less about why they don’t. We are proud of our ability to create questions and research studies that describe and explain consumer behaviours and activities, but we cannot force people to take part in marketing research because they don’t want to.

This question is brought to light by the widespread use of online access panels for research. Panels are made up of people who are willing and able to participate in research. We may have to repeat past mistakes concerning response rates if we don’t learn from them.

We need to study other disciplines to understand consumer behaviour and adapt our industry as needed to improve response rates.

You can slow down the decline in online response rates by putting in a lot of effort to improve the experience for respondents. Researchers can provide questionnaires to increase satisfaction by understanding the psychology of online interviews. Positive emotions must be encouraged in both the interview’s beginning and end. This can be done by using words and phrases that promote autonomy, competence and relatedness, and feelings of value. Researchers using online methods should be able to use these types of speech as a second language.

Key motivating factors

Psychology is “the study and application of psychology’s functions to affect behaviour in a particular context.” (Oxford English Dictionary). Marketing-related occupations are likely to have some knowledge about psychology’s history and the most famous psychologists, theories, and experiments. As we attempt to manipulate motivation to increase response and improve data quality, the maxim “a little knowledge can be dangerous” is more pertinent than ever.

Psychology is a relatively recent science that grew out of the wonderful philosophy tradition. It was established in the late 1800s by Wilhelm Wundt, who set up the first psychology labs in Leipzig in 1875. William James also founded the Harvard Psychology Labs in Harvard in the same year.

Harvard student James included Edward Thorndike, who later experimented with puzzle boxes and cats (1898). This led to the creation the “law of effect”, which states that stimuli that produce a pleasant or satisfying “state of affairs” are more likely to be repeated in that situation. In contrast, unpleasant, irritating, or uncomfortable responses are less likely to be repeated in the same situation.

Ivan Pavlov, a Russian physiologist and not a psychologist, studied the physiology behind digestion. He had been specifically interested in the salivation of dogs when food is present. His experiments with associative stimuli (ringing a bell together with food presentation and then ringing the bell alone resulting in salivation) led to the creation of conditioned reflexes. Although this work is well-known now, it was not widely recognized outside scientific circles until the 1920s.

As it was known, the Behaviourist school would continue to be a dominant force in the field for the first two-thirds of the 20th century.

BF Skinner’s work is the most prominent among the Behaviourists. His experiments were based on Pavlov’s ‘classical conditioning of stimulus-response and Thorndike’s law of effect. Skinner experimented on rats and pigeons using what is now known as the Skinner Box. Skinner Box is a collection of consequences applied to certain behaviours, such as pressing a lever. Skinner uses the term Operant Conditioning to describe the consequences of certain behaviours on future behaviour. There are five types of Operant Conditioning:

Positive reinforcement: A positive consequence is a behaviour reinforced by positive reinforcement.

Negative reinforcement: A behaviour reinforced by the consequences of ending an adverse condition.

Positive punishment is when a person’s behaviour is made worse by an adverse condition.

Negative punishment is when a person’s behaviour is made worse by suffering the consequences of losing a favourable condition.

Extinction is when a previously reinforced behaviour ceases to be effective, negative reinforcement is not provided, or punishment is applied.

Operational conditioning continues to be demonstrated in behaviourist experiments with animals and humans. Many of these experiments with human subjects would now be considered unethical. Albert B.’s case is perhaps the most well-known behaviourist experiment. Albert B. was born to Rosalie Rayner and John B. Watson, workers at the clinic. Albert was nine months old when he was first exposed to various stimuli. Albert was not afraid of a white rabbit, a rabbit or a dog. He also did not fear cotton wool, burning newspapers, and masks with or without hair. Albert was encouraged to fearfully strike a metal bar with a hammer, making a loud sound. Albert’s fear response was as expected from a child this age.

Albert was 11 months old when further experiments were conducted. Albert was presented with a white rat, and, as he tried to touch it, the metal bar was again struck. Albert was overcome with fear. Albert’s fear was exhibited when the white rats were introduced to him (without making a loud sound).

A few days later, Albert received the rat again, alone. He showed fear.

Albert began to experiment with various objects, including a rabbit, dog, fur coat and even Watson wearing a Santa Claus mask.

Albert B. was taken from the hospital without reconditioning (to remove conditioned responses). While speculation continues about his psychological future, James B. Watson has a distinguished career as an advertising executive at J Walter Thompson.

The Behaviourists’ primarily mechanical view was starkly opposed to the philosophy-based tradition and would have been difficult for Wundt to recognize.

Psychology’s Cognitive Approach introduced a more human aspect to its approach. Maslow’s Hierarchy of Needs, Herzberg’s Two Factor Theory, and Vroom’s Expectancy Theory brought back human needs and wants.

The Maslow Hierarchy, often presented as a pyramid, is the most well-known theory for non-psychologists. The theory states that people have needs that they want to satisfy. These needs are ranked in a hierarchy that starts with the physiological (food, water, air and sleep) and moves up to safety and security (structure order, security and predictability). Finally, there is love and belonging (friends, companions, a supportive group, identification with an intimate relationship), and finally, esteem (recognition by other people that results in feelings of prestige, acceptance, status and self-esteem that, result in feelings of adequacy and competence These are called “D-needs” (or “deficiency requirements). An individual who is unable to satisfy D-needs will feel anxious. However, once satisfied, they will feel nothing.

The need to self-actualize (personal growth and fulfilment) is at the top of the pyramid. These needs are not lacking.

Maslow’s hierarchy was revised in 1970 to include the desire “know and understand” and the “aesthetic” need. These needs were more important than self-actualization. This revision is often ignored in management theories and textbooks.

One could argue that the above description summarises the marketing researcher’s knowledge about psychological and human motivation.

What does this (Lack Of) Knowledge mean for the way we treat respondents

Marketing researchers, like other sciences, rarely refer to respondents as people. We call them respondents. Some psychologists refer to people as “organisms” while philosophers refer to them as “agents.” This would suggest that we fall more in the behaviourist camp than in the cognitive.

Behaviouralism would suggest that we must create a stimulus to get a response. The stimulus needs to be enforced on an ongoing basis to be effective. However, we have not been willing to offer incentives to survey participants based on the notion that “if they pay this time, we’ll have to pay every other time, and they will want more!” This is a strict behaviourist view of reality. Traditional research practitioners don’t care about the re-enforcing aspect and want a “fresh” sample each time they do a research project.

It is not possible to then operate within an operant conditioning model.

Research works by appealing to Maslow’s hierarchy’s esteem needs. Interviewers ask respondents for help with the “important” survey, and they inform them that they have been “chosen.” Marketing research associations have promoted the industry as “your opinion counts”, and political opinion polls are widely reported. Respondents are expected to be given “esteem” in return for their opinions.

The internet and other media have helped dispel the myth of low self-esteem. The public is now aware that marketing research is done for commercial purposes and our clients spend billions of dollars with us. Very little of this money ends up in the hands and efforts of those who give their time to provide us with knowledge.

Participants in market research surveys may not need to be able to accept a vague promise of self-esteem if they are to understand this reality. As an industry, we have tried to convince respondents that they might learn something about themself by participating in market research surveys – appealing to higher levels of ‘self-actualization’. However, I believe that a promise to esteem is often empty.

We will not be able to reverse the decline in response rates if we view the world only in psychological terms.

Psychology is changing.

Festinger’s Theory of Cognitive Dissonance (1956) allows us to understand the issues around choices and how they are made individually. Each person interviewed for an interview has the freedom to choose whether or not they want to participate.

Dissonance refers to psychological anxiety. It is an unpleasant state of mind. Rational people will try to reduce dissonance. Dissonance occurs when two cognitions (elements of knowledge) are relevant and opposite/contradictory to each other. A smoker might enjoy smoking but be aware of the health risks associated with it.

You can reduce dissonance by either decreasing the importance of conflicting beliefs or adopting new beliefs that balance the situation. The dissonant smoker may believe that smoking suppresses appetite and that being overweight is more dangerous than smoking.

He may also dismiss some health warnings as unscientific or point to a 95-year-old smoker as an example.

Most researchers will be familiar with the Free Choice paradigm, a model of the theory. This paradigm states that dissonance will result once you make a free choice. All negative cognitions about the selected object will conflict with all positive cognitions. The best way to reduce dissonance is to increase the positive cognitions regarding the chosen object while making the rejected ones less favourable. This is what we call “Post Purchase Justification”, and it will be familiar to anyone who works in Automotive research.

J. Brehm (1956) conducted an experiment to test Free Choice. The experiment was conducted under the pretence of market research, perhaps the first instance of “psyugging”. Homemakers were then asked to rate various appliances on their desirability. Then, they were asked to pick an appliance for their use. After making a choice, the desirability scores were again taken. The appliance chosen was rated higher than the one it had been previously, while the ones rejected were rated lower. The effect was stronger in areas where the appliances were rated more highly and closely (i.e., it was more difficult to choose) than if they were closer together. This is an excellent example of post-purchase justification. It also explains why this rarely happens with buying a box of matches.

Although it purports to be, cognitive Dissonance Theory is not a comprehensive theory of motivation or behaviour. People are seen as almost moving from one choice to another rather than towards a higher or future goal.

Self-Determination Theory is one of the most recent theories. It seeks to provide a more generalized approach. Richard M. Ryan of the University of Rochester is the leading proponent of SDT. According to the theory, people are “active organisms with innate tendencies towards psychological growth and development.” They “require ongoing nutrition and support from the social environment to function effectively.” This means that the social context can either support the natural tendency toward active engagement or hinder it.

SDT differentiates between different motivations based on the reasons or goals that lead to action. The most crucial distinction is between intrinsic and extrinsic motives when something is done because it is exciting or enjoyable.

Intuition-based motivation is characterized by creativity, high quality, and enjoyment.

It is important to note that extrinsic motivators can include any particular outcome, such as peer approval or selfless acts. These are often called intrinsic motivation in marketing research.

The task, the person and the social context all play a role in motivation. Motivation is not a single thing. The same task may have different motivation levels for each person.

SDT views motivation as a continuum, which is neither hierarchical nor uni-directional.

As motivation increases, performance on a task can improve (both in terms of output and well-being).

The critical point of the continuum is the transition between Identification and Introjection. The step between Introjection and Identification is where the person recognizes the importance of behaviour and accepts it as their own. Extrinsic motivation’s final stage (Integration) shares many characteristics with Intrinsic motivation. Both are independent, and there is no conflict (dissonance).

The person’s current situation and prior experience will determine their position on the continuum for a given task. The task is not the only thing that affects movement along the continuum. It’s more about how you perceive competence (“I did that task very well”) and externalizing or internalizing causality (“They are making me do it”).

SDT research has revealed that rewards can be detrimental to intrinsic motivation. This is because autonomy is compromised. This is contrary to what behaviourist psychology would predict. This effect is more evident when linked to an individual’s performance.

The impact of new psychology in Market Research

Cognitive Dissonance theory helps us understand how people decide whether or not to take part in the survey. Recognizing that people know (cognition) of marketing research and related cognitions allows us to try to reduce dissonance when we send out invitations or personal approaches. We need to make sure that research experiences align with, or better than, pre-existing cognitions.

SDT can have interesting implications on how research should be conducted before and after interviews. To “internalize regulation”, we know that autonomy must be encouraged. To encourage respondents to participate in research on their initiative, it is essential to foster feelings of autonomy. It is almost impossible to avoid discomfort when the end of the survey process is dominated by a thank you and an “if there are any doubts, please call”. We may need to learn how to give feedback on survey performance and the correct language.

Research is (probably) positioned at the continuum’s External Regulation or Introjection points. However, there is evidence (Comley & others) that people report doing surveys because they like to give their opinions. It is unclear if this is a self-serving reevaluation of cognition and not a firm belief. We approach respondents by offering a reward, often small or subtly pressing them into participating. This is done using their sense of altruism (Pro-Social behaviour) or politeness(Politeness Theory).

It is time to stop treating respondents as an inexhaustible resource that can be plucked for our purposes without regard for the industry’s future. To re-engage respondents, we must come up with new methods. These techniques should be able to meet both the research industry’s needs and the needs of the respondents.

This is why it matters so much to an online access panel company

We are an online access panel provider responsible for providing respondents with surveys. Although we don’t have any control over the content of the survey, the reality is that respondents are exposed to it. However, interviews must be conducted.

We are interested in strengthening (research) behaviour, not traditional research methods. The industry is reducing revenues per interview due to a finite resources. We can motivate our respondents to be more productive in recruiting and retaining them.

The knee-jerk reaction (behaviourist?) The knee-jerk (behaviourist?) reaction is to offer monetary incentives to participate in surveys. SDT warns that such rewards may cause a decline in motivation. Cognitive dissonance states that a small amount of reward can cause dissonance.

Suppose we can motivate people to intrinsic motivation by moving them along the SDT motivation continuum. There should be higher response rates to surveys (which would mean less email invitation intrusion) and more enjoyment, effort and creativity from the panellist leading to better quality research.

Through the social aspects of our panel and the shared webspace, we already strive to foster autonomy and competence. We know that there is more we can do in this area. The motivational state is the essential information that’s missing.

The latest internal research aims to better understand panellist motivation, with the ultimate goal of creating an incentive program that will foster individual motivation.

The result of our study

First, it is essential to remember that this research project has a response rate. Those who respond to the survey will be motivated to do so. Those who don’t will not. This should not be an issue since we aren’t trying to divide the respondents into different motivational types based on their prior knowledge.

The invitation was as appealing as possible, and it appealed to respondents’ senses of community and the importance and research. Respondents will also be shaping the future direction of all panel members. This made the research “value” paramount. We were able to keep the survey open longer than usual to ensure the highest response rate because we didn’t have to meet deadlines. There were 1,423 interviews conducted across all age groups and for both genders.

The research was conducted in the classic NPD demand study format. Four future reward schemes and the existing charity and prize drawing mix were evaluated. These four cash-based schemes were:

  • Redeemable vouchers with points
  • You earn points toward cash rewards
  • Spend points in our online shop to earn points
  • Actual cash payments per survey
  • Each interview was given the same value, and it increased with increasing length.

There were four interview lengths tested. Respondents were randomly assigned to a group according to their interview length.

Respondents in simulation first decreased their consideration by refusing cash rewards they did not find appealing. The remaining invitation/reward types were then presented and ranked according to preference. The respondent had the option to pick “none” at any stage of the ranking process. Respondents were not required to complete the survey in return for any incentives.

OpinionWorld panels at SSI currently make donations on behalf of respondents to charities and offer sweepstakes for cash or prizes. The panel is not affected by not offering cash payments one-on-one in terms of panel response rates. Respondents might actively refuse cash or cash proxy such as points. This was not the case. Nearly everyone, except a small minority, admitted to being attracted to one or more cash rewards in the survey.

The most requested option was a cash payment (62%), with points for cash (52%) being the second most sought-after option. These options attracted 82% of respondents.

If all five options were applied to the panel, the expected uptake would have been:

  • Pure cash 40%
  • Cash points 27%
  • Charity/prize draw 13%
  • Points redeemable for vouchers 13%
  • Shop online 7%

The two above analyses show that no one size fits all and that respondents do not desire cash rewards for participating in surveys. This is not surprising as many panellist motivation surveys have shown that enjoyment from participating in surveys is a critical factor for their decision to join panels. It is not surprising that respondents prefer cash to pseudo-cash. According to classic economics, cash is the best way to give people what they want.

The saliency of rewards was tested using questions asking if the reward system needed to be changed (in advance to asking about possible changes, of course) and asking how rewards are considered before taking surveys. Over half (58%) did not believe that any changes were necessary. When this question was analyzed concerning Intrinsic motivation, there were significant differences. Only 39% of those with high intrinsic motivation wanted to see change. The least motivated people were the most demanding of change, with 52 per cent wanting to see change. Only 30% of respondents to the second question said they “often” or “often” think about the rewards that would be offered before deciding to participate. Although there is an intense desire for change, it is not universal. Actual rewards are secondary to whether or not to take part in a survey. A second question in the survey asked about reasons people don’t participate in surveys. It found that “too busy (21%), as well as the survey’s end-by date (26%), were the most common reasons. Only one in ten mentioned low rewards.

Only the length of an interview will affect the choice of the reward mechanism. People will switch from cashpoints to real cash as the interview length increases. Other mechanisms were stable at all lengths of interviews.

We observed how cooperation rates could change if there were different reward options. Respondents were allowed to choose “none” in any scenario. Although interview length plays a part in reducing participation, it wasn’t marked until the end of the 35-minute survey. Around 17% of respondents said no to the survey invitation if they received at least one reward scheme. Although this is not a prediction of an 80% plus response rate, it can be interpreted as an indication that there could be some improvement in response rates if we offer more reward schemes.

It is almost certain that a range of reward programs will be required, not just one “most preferred” scheme. To determine their likely responses to two scenarios, we asked people who would like a change. Just under half of the respondents (49%) stated that they would not do as many surveys if the reward system were changed to something they did not like. However, 83% of respondents said they would take more surveys if the reward scheme were changed to something they liked.

The research focused on the respondents’ motivational state in an SDT context. This was a critical question. We were particularly interested in testing the possible adverse effects of individual rewards or one-on-one payments on intrinsic motivation since we don’t currently offer these rewards.

SDT predicts that intrinsic motivation will be determined by how you feel about the task’s “value”. SDT also predicts that intrinsic motivation will be positively correlated with autonomy measures in completing the task.

We presented the panellist with an IMI (intrinsic motivation inventory) standard questionnaire at the end of the research. This set of questions covers interest/enjoyment and value/usefulness, and perceived choices while performing an activity (completing a survey). As one would expect from an internet survey, the measure of perceived choice (autonomy) scored high. However, it is positively correlated to intrinsic motivation.

Autonomy VS Intrinsic Motivation

More clearly, intrinsic motivation is correlated with the measure of value/usefulness and scores across the entire range.

Value VS intrinsic motivation

This finding has great potential to increase intrinsic motivation. It can demonstrate the value of research to oneself or others. The research’s uses could be reflected in feedback, increasing intrinsic motivation and feelings of value. Any promise of utility must be supported by reality. This includes autonomy, support, and, where appropriate, competence and relatedness. The survey instruments must be sufficient in quality not to undermine any notion of utility.

After attempting to instil a sense of competence and value through the questionnaire, a random half-selected sample was given an extrinsic reward: an entry into that month’s prize draw. To begin the survey, no incentive was offered. The IMI question batteries were then sent to the respondents. SDT theory predicts that the rewarded group is less intrinsically motivated than that which received no reward. As this simple analysis shows, it was true.

These movements are statistically significant, even though they are relatively small. The sentence “For doing this survey, we will allow you to enter the monthly prize draw” was added to this paragraph.

Thank you for taking the time to complete this essential survey. We understand that some surveys are challenging and that this one was not easy.

We will consider your opinions as we decide whether to change the rewards program.” This finding clearly shows the need to have a better understanding of the psychology behind survey-taking and a greater appreciation for what researchers may be doing to respondents through our questionnaire designs and survey rewards.


We cannot allow response rates for marketing research to drop further. Otherwise, we risk losing our industry. Online access panels are not the panacea for online research. In the short time that online research was viable, we have seen a reduction in responses that matches what the entire industry has taken over 50 years.

Understanding the psychology of interview situations can help us identify areas where we might have gone wrong. Of particular importance are the new theories of motivation. These theories already indicate areas in which we can improve respondents’ experience without having to experiment. Notably, the first approach. In a new way, the end of the interview might be more positive. These principles will help you to be more motivated in marketing research.

Daniel Williams

Daniel Williams

Dan is from Oxford, UK and is one of our reviewers who’s been with HuginX the longest. With a background in membership support at one of the largest survey companies, Dan is in an excellent position to review panel communities.

Compare panel features
  • Total (0)