The Middle Way is Great for Contentment, Less Great for Customer Surveys

by Tom Hagy for HappyOrNot

The HappyOrNot’s Smiley Terminals famously sport four bright beautiful buttons. Just looking at them can inspire joy. Some might think the four-button platform was a random choice. Others might think more buttons would better gauge customer experience by depicting a larger swath of the emotional spectrum. Still others think giving customers a middle button would be easier on them – they may not want to struggle with their emotions or they just might not want to think that hard.

First and foremost – the four-button format is deliberate!  Believe it or not it has been the subject of numerous studies over the years. In the end, we arrived at our magic number for the way our terminals are applied, taking into account the need to:

• Offer enough selections to accurately gauge customer emotion.
• Make giving feedback quick and easy for the customer.
• Capture feedback immediately while the customer experience is fresh.
• Get the customer to come down on one side or the other.

Siddhartha Gautama, more commonly known as The Buddha, taught that the way to achieve contentment in life, or Nirvana, is the Middle Way. He guided his followers to embrace neither unbridled self-indulgence nor extreme self-denial. Whether people followed these teachings or came to the same conclusion on their own – comically and tragically, perhaps, through trial and error – this approach has served many millions of people quite well for thousands of years.

Many folks feel this way about life in general. They embrace moderation and seek compromise. Whether it is their political selections or how often they eat ice cream, they head for center. And when we don’t migrate toward moderation the consequences can range from socio-economic upheaval to difficulty fitting into our pants. Occasionally we do something wild and crazy. We sound off with extreme statements just for the fun of it.  Sometimes we make life-altering decisions with little thought or lucidity (fueling demand in Las Vegas for drive-through wedding chapels). But for most of us that’s the exception.

Everything is OK. Now What?

When it comes to surveying customer experience, though, the middle choice may not be our enemy, but it certainly isn’t our friend. Making business decisions and spending money on research demands more than “everything was fine” or “OK.”  If anyone said you were an “OK” kisser you would be left to ponder the statement’s meaning. Was it you? Was it them? What’s the remedy? Additional opinions? Practice? Dental hygiene? Relocation?

Likewise, what is an operations manager or marketer supposed to do with “fine”?  If “Choice A” means terrible and “Choice C” means incredible, what good is “Choice B”?  And what is a customer to do with 26 choices ranging from “A” to “Z”? Who has the time?

Identification and Mitigation

It’s important in the survey game to identify bias trends and mitigate those trends with an appropriate number of choices in order to capture meaningful data. But, regardless of the methodology used, people will head for the middle. Called by many names, like “middle bias,” “centrality preference,” or “the center stage effect,” this tendency is one that gets the most attention when looking at survey biases.

What number of choices is the right number to give us actionable data? What number will wear the customer out or scare them away? Should survey respondents be given the choice of saying everything was OK, or should we make them come down on one side of the big ratings divide or the other? That’s what this article will attempt to answer, and explain our devotion to the number four.  

The Numbers Game

Before we get into the middle choice, let’s take care of how many choices is the right number of choices.

One of the more frequently cited papers on the subject comes from Duane F. Alwin, a sociology and demography professor at Pennsylvania State University. In his January 1992 article – Information Transmission in the Survey Interview: Number of Response Categories and the Reliability of Attitude Measurement — published in the journal Sociological Methodology, Alwin took an educated stab at the right number.

He attempted to find a balance between many choices, which he said increases the reliability of the survey (but at a decreasing rate) and the rate of response. Alwin concluded that, for his money, the magic number falls between four and seven.  

Another study concluded that a greater number of choices – e.g. from 10 to 101 — will increase reliability and allow the respondents to share their emotions more accurately. But, researchers Carolyn C. Preston and Andrew M. Colman at the University of Leicester said, the better numbers are 7, 9, and 10.  Their study was titled Optimal Number of Response Categories in Rating Scales: Reliability, Validity, Discriminating Power, and Respondent Preferences and was published in Acta Psychologica in 2000.

Preston and Colman continued, though, the purpose and circumstances of a survey – such as time pressure – must be considered “in order to prevent the respondents from becoming frustrated and demotivated.” Uses of three- or five-point scales are “likely to be perceived by the respondents as relatively quick and easy to use.” What you may lose in extreme and, arguably unnecessary, precision, you gain in participation.

The Middle Choice: Pro and Con

In March 2018 marketer and online learning professional, Kent Hendricks, wrote an article titled The Center Stage Effect: This is Why You Choose The Product in the Middle (Even if It’s Bad). In some cases, he said these choices come with little risk, like which way to drive home or which pen to use. This tendency also influences, though, which products consumers buy or which payment plans they select.

“When choosing between three packs of chewing gum, customers chose the middle option half the time, while 29.17% chose the left option and only 20.83% chose the right option,” Hendricks wrote.  It’s easy to see the commercial applications of these findings especially when selling thousands or millions of units. For example, he wrote, retailers found that simply moving toothbrushes from the top shelf to the middle shelf led to an 8% increase in sales. (Toothbrushes, unlike liquor, aren’t categorized as top-shelf or bottom-shelf, but then we go back to looking at where bottles sit on a horizontal line and, once again, the middle calls us.)

When asking participants to take a survey there are three key factors to consider, according to Hendricks: 1) cultural bias; 2) memory; and 3) the “gaze cascade effect.”

Cultural bias. A participant’s background and experiences influences their selections. Studies show that a person from the United States is more likely to pick an option that is leftmost on a horizontal scale since they read left to right. Hypothetically then, a person who speaks Hebrew or Arabic is more likely to pick the rightmost option. In Western culture, we place important people, like the CEO or the contest winner, in the middle. Because we humans rely so heavily on social constructs to get along we avoid deviation and, when taking a survey, head for the safety of the middle.

Memory. If an experience wasn’t particularly memorable, the participant might rate it as neutral. Additionally, when participants in a study were asked to recall items along a horizontal line, the middle and leftmost items were the most attended to and remembered.  Capturing a person’s opinion before they forget the details of their experience greatly improves accuracy.

The gaze cascade effect. This refers to brain and eye movement communication. The human’s brain moves much faster, about .5 milliseconds faster, than your eyes. When presented a horizontal set, your brain has already captured the entire scene. By the time your eyes reach the end of the set your brain is already cozy in the middle.

“Researchers have discovered a strong correlation between what you look at and what you choose. When you make a choice, you will gaze – either intentionally or unknowingly – at the item you eventually select. The item that gets the most attention gets chosen. You look more often at items you like,” Hendricks wrote. “But the reverse is also true. You like items you look at. This creates a feedback loop: looking at something causes you to like it, and liking it causes you to look at it more often.”

Middle Masks Reality

George F. Bishop with the University of Cincinnati, in June 1987 wrote in Public Opinion Quarterly, that “people are significantly more likely to select the middle response alternative on an issue when it is explicitly offered to them as part of the question than when it is omitted.” Even the mere mention of a middle choice when asking the question “makes it more likely that respondents will select it, even though it is not offered to them as an explicit choice.”  

“People who select a middle response alternative when it is offered would not necessarily answer the question in the same way as other respondents if forced to choose sides on the issue,” Bishop determined. That means the middle choice can mask the customer’s true feelings.

The “centrality preference” is said to have first been discovered and studied by Nicholas Christenfeld at the University of California, San Diego. His Choices from Identical Options was published in many journals including the American Psychological Society’s Psychological Science. Based on six studies of choices, whether it was checking a survey box or choosing a product from a grocery shelf – Christenfeld said people “avoided the ends and tended to make their selection from the middle.”

As examples, he said 71% of shoppers picked items from the two middle shelves. Other types of choices came out differently – like which route to take on a map – in which most people avoided middle routes. Christenfeld speculated that sometimes these choices are intended to minimize effort. “Aiming for the middle of a group may require less concentration than going for extremes, and not worrying about where to turn until one has to turn may keep one’s mind clear of unnecessary information.”

“It’s hard enough deciding what cereal to buy without deciding which particular box, and hard enough deciding where to go without deciding how to get there,” Christenfeld concluded.

“If item location influences preference during the millions of purchasing choices that occur every day, it will be exerting a substantial influence on consumer behaviour,” concluded researchers Paul Rodway, A. Schepman and J. Lambert in a 2012 paper published by The British Psychological Society.

Their research was built upon that shared by The British Psychological Society in 2006, which found that people overestimate the performance of people who sit in the middle. Priya Raghubir and Ana Valenzuela examined 20 episodes of the American television show “The Weakest Link” in which contestants in a semi-circle vote others off the show, sending them home crushed by the experience. The players in the two central positions won 45% of the games. Contestants stuck at the ends won just one out of 10.

So back to our questions: How many choices should a survey offer and should there be middle?  For the following reasons HappyOrNot goes with four buttons.

People will:

• Pick the middle even if their true feelings lean negative or positive.
• Avoid the strain of too many choices.
• Avoid being harsh or deviant, so need some help being frank.
• Participate in a very quick survey so they can get on with their day.

Managers and Marketers want:

• An adequate array of choices so they receive accurate and actionable information.
• A sufficient number of responses so they can spot genuine blips and trends.
• Truthful feedback when customer experience is fresh in their minds.
• Timely feedback so they can act in real time if necessary.

If you’re still not convinced, how about this:  If serious researchers were dedicated enough to watch and analyze 20 episodes of an American game show, aren’t their conclusions worth considering?