Measuring Attitudes: Thurstone and Likert Scales Explained

Imagine you’re trying to understand how your community feels about a new health policy, or whether people truly believe in the importance of balanced nutrition. How do you measure something as intangible as an attitude? This is where attitude scales come in-powerful research tools that transform abstract feelings into measurable data. From public health surveys to educational research, these scales help us quantify what people think, feel, and believe about the world around them.

What are attitude scales and why do they matter? 🔗

An attitude is more than just an opinion-it’s a psychological tendency to evaluate something with a degree of favor or disfavor. Whether we’re assessing feelings toward a healthy lifestyle, attitudes about food safety, or perceptions of nutrition education programs, we need reliable ways to measure these evaluations. Attitude scales serve as instruments specifically designed to capture both the intensity and direction of an individual’s attitude toward a particular concept or object.

These scales don’t just tell us whether someone agrees or disagrees with something. They reveal the strength of that agreement, the nuances of opinion, and patterns across populations. This information becomes invaluable when designing public health interventions, evaluating nutrition education programs, or understanding community needs.

Thurstone’s equal-appearing intervals: The pioneer approach 🔗

In the 1920s, psychologist Louis Leon Thurstone revolutionized attitude measurement by developing a systematic method for creating scales with mathematically equal intervals. The method of equal-appearing intervals was his most practical contribution, though he actually invented three different scaling methods.

How Thurstone scales are built 🔗

Creating a Thurstone scale is like building a measuring stick for opinions. First, researchers generate a large pool of statements-typically 80 to 100-that represent different positions on a topic. For example, if measuring attitudes toward plant-based diets, statements might range from “Plant-based diets are nutritionally inadequate” to “Plant-based diets are the healthiest option for everyone.”

Next comes the critical step: a panel of judges rates each statement on an 11-point scale based on how favorable or unfavorable it is toward the concept being measured. Notice that judges aren’t sharing their personal opinions-they’re objectively evaluating where each statement falls on the favorability spectrum. Think of them as calibrating the measuring stick rather than using it.

The statistical magic happens when researchers calculate the median rating for each statement, along with the interquartile range (a measure of how much judges agreed with each other). Statements are then selected at equal intervals across the scale, with preference given to those with the smallest interquartile range-meaning judges most consistently agreed on their placement.

Using a Thurstone scale in practice 🔗

Once the final scale is constructed, respondents simply check “agree” or “disagree” for each statement. Their attitude score is calculated by averaging the scale values of all statements they agreed with. Someone who agrees only with negatively-weighted statements receives a lower score than someone who agrees with positively-weighted ones. It’s elegant in its simplicity, yet powerful in its precision.

The beauty of this method lies in its attempt to create true interval-level measurement-where the distance between a score of 3 and 4 is theoretically the same as the distance between 7 and 8. This makes statistical analysis more robust and meaningful.

Likert’s summated ratings: Simplifying attitude measurement 🔗

While Thurstone’s method was groundbreaking, it was also time-consuming and complex. In 1932, Rensis Likert introduced a simpler yet equally powerful approach that has become the most widely used attitude measurement technique in social science research. The Likert scale streamlined the process while maintaining reliability.

The structure of Likert scales 🔗

A Likert scale consists of multiple items-typically statements-that respondents rate on a symmetric scale. The classic format uses five response options: strongly disagree, disagree, neither agree nor disagree, agree, and strongly agree. Each response is assigned a numerical value (usually 1 through 5), and these values are summed across all items to create a total score-hence why it’s also called a summated scale.

For instance, a nutrition attitude scale might include statements like “I believe eating breakfast is essential for health” or “Planning meals ahead saves me time and money.” Respondents indicate their level of agreement with each, and researchers sum the responses to get an overall attitude score. Higher total scores typically indicate more positive attitudes toward the concept being measured.

Why Likert scales became so popular 🔗

Unlike Thurstone scales, Likert scales don’t require a panel of judges to pre-rate items. This makes them much faster and less expensive to develop. They’re also more flexible-researchers can use 5-point, 7-point, or even 4-point or 6-point scales (forcing respondents to lean toward agreement or disagreement by eliminating the neutral option).

The method assumes that the underlying phenomenon being measured is continuous and that responses across multiple related items will collectively reveal a person’s true attitude. Think of it as triangulating on someone’s feelings by asking the same question in slightly different ways. If someone consistently agrees with pro-nutrition statements and disagrees with anti-nutrition ones, we can be fairly confident about their overall attitude.

Key differences from Thurstone’s approach 🔗

While both methods measure attitudes, they differ in important ways. Thurstone scales ask respondents simply to agree or disagree with statements, while Likert scales capture degrees of agreement. Thurstone scales rely on judges to assign scale values, while Likert scales let respondents’ patterns of agreement create the measurement. Thurstone aimed for equal intervals through careful construction, while Likert accepted that intervals might not be perfectly equal but that the summated score would still meaningfully reflect attitudes.

Real-world applications in public opinion surveys 🔗

Both Thurstone and Likert scales have found extensive use in measuring public opinion, particularly in areas that impact policy and health education. Government agencies use these tools to gauge public support for new regulations, understand community health priorities, and evaluate the effectiveness of public awareness campaigns.

In nutrition and public health contexts, attitude scales help researchers understand barriers to healthy eating, assess the impact of nutrition education programs, and identify target audiences for interventions. For example, a public health department might use a Likert-style survey to measure community attitudes toward school lunch programs before proposing changes. If scores reveal widespread concern about nutrition quality, policymakers have data-driven justification for budget increases or menu reforms.

These scales also prove valuable in evaluating behavior change interventions. When a community implements a farmers market nutrition program, pre- and post-program attitude measurements can reveal whether the intervention changed perceptions about fresh food accessibility, cooking confidence, or the value of local produce. Such evidence helps secure continued funding and informs program improvements.

Understanding the limitations: What attitude scales can’t do 🔗

Despite their utility, attitude scales come with important limitations that researchers must acknowledge and, when possible, address.

Social desirability bias: The polite answer problem 🔗

One of the most pervasive issues is social desirability bias-the tendency for respondents to provide answers they believe are socially acceptable rather than their true feelings. When asked about nutrition attitudes, people might overstate their commitment to healthy eating or underreport consumption of foods they know are “bad.” This is particularly problematic for sensitive topics like body image, dieting behaviors, or attitudes toward obesity.

Researchers have developed various strategies to minimize this bias, including ensuring anonymity, using indirect questioning techniques, and incorporating social desirability scales (like the Marlowe-Crowne scale) to statistically control for this tendency. However, no method completely eliminates the problem, and researchers must interpret results with this limitation in mind.

Acquiescence bias and response patterns 🔗

Some respondents tend to agree with statements regardless of content-a phenomenon called acquiescence bias. Others gravitate toward extreme responses or middle-of-the-road answers. These response styles can distort results, making scores reflect personality traits or cultural norms rather than true attitudes toward the topic being studied.

The abstraction challenge 🔗

Attitude scales measure what people say about their beliefs, not necessarily what they actually believe or how they’ll behave. There’s often a gap between stated attitudes and real-world actions. Someone might strongly agree that “eating vegetables daily is important” while rarely actually doing so. Attitudes are abstract constructs, and reducing them to numbers inevitably loses some complexity and context.

Cultural and contextual factors 🔗

Attitudes vary significantly across cultures and contexts. A Likert scale developed in one cultural setting may not capture the nuances of attitudes in another. The meaning of “neither agree nor disagree,” for instance, might represent genuine ambivalence in one culture but politeness or social harmony in another. Cross-cultural comparisons require careful consideration of these differences.

Limited depth and nuance 🔗

While attitude scales excel at quantifying overall sentiment, they may miss the rich, multifaceted nature of human attitudes. They tell us the “what” and “how much” but often not the “why.” This is why many researchers combine scales with qualitative methods like interviews or focus groups to gain deeper understanding alongside quantitative measurement.

What do you think? Have you ever filled out a survey and found yourself struggling to express your true feelings within the limited options provided? How might combining quantitative attitude scales with qualitative research methods give us a more complete picture of people’s beliefs and motivations?

References

1. https://socialwork.institute/research/attitude-scales-research-measuring-opinions-beliefs/
2. https://conjointly.com/kb/thurstone-scaling/
3. https://www.statisticshowto.com/thurstone-scale/
4. https://en.wikipedia.org/wiki/Likert_scale
5. https://conjointly.com/kb/likert-scaling/
6. https://en.wikipedia.org/wiki/Social-desirability_bias
7. https://www.sciencedirect.com/topics/psychology/social-desirability-bias