Attitude Scales in Research: Measuring Opinions and Beliefs

When researchers need to understand what people think or feel about social issues, they face a unique challenge: how do you measure something as subjective as an attitude? Attitude scales provide a solution by transforming opinions, beliefs, and feelings into quantifiable data. These measurement tools have become essential in social work research, helping professionals understand public sentiment on everything from mental health stigma to community support programs.

What are attitude scales? 🔗

Attitude scales are specialized instruments designed to measure and quantify people’s opinions and attitudes in research settings. At their core, attitudes represent mental states or emotional responses that lead individuals to evaluate something in either a positive or negative light. These evaluations can involve people, ideas, policies, or social phenomena.

The purpose of these scales extends beyond simple yes-or-no questions. They capture both the intensity and direction of someone’s attitude, providing researchers with nuanced data about how strongly people feel about particular issues. In social work, this might mean measuring attitudes toward foster care systems, substance abuse treatment programs, or immigration policies.

Two primary methods have dominated attitude measurement since the early 20th century: Thurstone’s method of equal-appearing intervals and Likert’s summated ratings. Each offers distinct advantages for different research situations.

Thurstone’s method of equal-appearing intervals 🔗

Developed by Louis Leon Thurstone in 1928, this scale was the first formal technique to measure attitudes systematically. Originally created to assess attitudes toward religion, the Thurstone scale has since been applied to diverse topics across psychology and sociology.

How Thurstone scales work 🔗

The Thurstone method involves a two-stage process. First, researchers develop a large collection of statements about a particular topic, ranging from highly favorable to highly unfavorable. A panel of judges then rates each statement on a scale, typically from 1 to 11, based on how favorable or unfavorable it appears. The judges are not expressing their own opinions but rather evaluating how each statement reflects different attitude positions.

After this rating process, researchers select statements that are evenly distributed across the scale and show strong agreement among judges. The final scale presents respondents with these pre-weighted statements, asking them to indicate which ones they agree with. The respondent’s attitude score is calculated as the average of the scale values for all statements they endorsed.

Strengths and limitations 🔗

The Thurstone scale offers important advantages. It provides interval-level measurement properties, meaning the difference between scores of 3 and 4 is mathematically equivalent to the difference between 7 and 8. This precision allows for more sophisticated statistical analysis compared to simpler scales.

However, the method has notable drawbacks. Developing a Thurstone scale is time-consuming and expensive, requiring extensive effort to gather statements, recruit judges, calculate median values, and finalize the instrument. Additionally, the values assigned by judges may inadvertently reflect their own attitudes, introducing potential bias into the scale construction.

Likert’s summated ratings 🔗

The Likert scale, developed by Rensis Likert in 1932, emerged as a simpler alternative to the Thurstone method. It has become the most widely used approach to measuring attitudes in survey research today.

Understanding Likert scales 🔗

Unlike Thurstone scales, Likert scales ask respondents to indicate their level of agreement or disagreement with a series of statements. The most common format uses five response options: strongly disagree, disagree, neither agree nor disagree, agree, and strongly agree. Some researchers prefer four or seven-point scales, with even-numbered scales forcing respondents to lean toward agreement or disagreement by removing the neutral option.

Each response is assigned a numerical value, and the final score is the sum of ratings across all items, which is why these are sometimes called summated scales. This approach assumes that all items measure the same underlying attitude and that distances between response categories are roughly equal.

Why Likert scales gained popularity 🔗

The Likert method offers several practical advantages over the Thurstone approach. It does not require a panel of judges to assign scale values, making it faster and less expensive to develop. Research has shown that Likert scales produce reliability coefficients comparable to or higher than Thurstone scales, despite their simpler construction process.

The use of multiple items rather than single questions also improves measurement quality. Multiple indicators help reduce random measurement error and provide broader coverage of the attitude being measured. A well-designed Likert scale typically includes six to thirty statements, mixing positively and negatively worded items to minimize response bias.

Applications in social work research 🔗

Attitude scales serve numerous purposes in social work research and practice. They are particularly valuable for measuring opinions about social programs, gauging attitudes toward vulnerable populations, and assessing public sentiment on policy issues.

In opinion surveys, researchers use these scales to produce quantifiable data about the strength of respondents’ beliefs and feelings. For example, a social worker might use a Likert scale to measure community attitudes toward establishing a homeless shelter or implementing needle exchange programs.

These scales also help predict behavioral trends. Customer satisfaction data can forecast future purchasing patterns, while employee engagement scores can indicate potential staff turnover. In social work settings, attitude measurements might predict volunteer participation, donation behavior, or community support for new initiatives.

Organizations use attitude scales to evaluate program effectiveness and identify areas for improvement. By measuring attitudes before and after interventions, researchers can assess whether educational campaigns, training programs, or policy changes have shifted public opinion in desired directions.

Navigating challenges in attitude measurement 🔗

Despite their widespread use, attitude scales face several methodological challenges that researchers must address.

Social desirability bias 🔗

One of the most persistent problems in attitude measurement is social desirability bias, where respondents provide answers they believe are socially acceptable rather than expressing their true feelings. This is especially problematic when measuring attitudes toward sensitive topics like racism, substance abuse, or stigmatized populations.

Research on substance users found that social desirability bias was associated with key health measures including recent drug use and stigma perceptions. Respondents may underreport socially undesirable attitudes and over-report more favorable attributes, leading to skewed results.

Researchers have developed several strategies to minimize this bias. Anonymous survey administration tends to elicit higher reporting of items affected by social desirability bias compared to in-person or phone-based data collection. Carefully wording questions and prefacing surveys with assurances of confidentiality can also help reduce these effects.

Acquiescence bias 🔗

Acquiescence bias occurs when respondents tend to agree with statements regardless of content, simply because they prefer selecting positive responses. This effect is particularly strong among children, elderly individuals, and people in institutional settings who may feel pressure to please the researcher.

Designing scales with balanced keying helps address this issue. By including equal numbers of positively and negatively worded statements, researchers can detect and correct for acquiescence bias, as agreement with contradictory items indicates response patterns rather than genuine attitudes.

Other measurement limitations 🔗

Attitude scales may not capture the full complexity of human opinions. Cultural and contextual factors can significantly influence how people interpret and respond to statements, making cross-cultural comparisons challenging. What seems neutral or balanced to researchers in one culture may carry different connotations in another.

Additionally, attitudes measured at a single point in time may not reflect stable, long-term beliefs. People’s opinions can fluctuate based on recent events, current mood, or temporary circumstances. Single items are notoriously unreliable, which is why researchers prefer multi-item scales despite the additional time required for respondents to complete them.

What do you think? How might researchers better account for cultural differences when designing attitude scales for diverse populations? In what situations might qualitative methods provide insights that attitude scales cannot capture?

References

1. https://en.wikipedia.org/wiki/Thurstone_scale
2. https://www.formpl.us/blog/thurstone-scale-guide
3. https://en.wikipedia.org/wiki/Likert_scale
4. https://conjointly.com/kb/likert-scaling/
5. http://www.jamescmccroskey.com/publications/25.htm
6. https://en.wikipedia.org/wiki/Social-desirability_bias
7. https://pmc.ncbi.nlm.nih.gov/articles/PMC5519338/
8. https://animalcharityevaluators.org/for-charities/develop-a-survey/social-desirability-scale/