In surveys, response scales play a crucial role in converting abstract concepts that are difficult to observe directly—such as attitudes, opinions, preferences, and satisfaction—into measurable data. The quality and reliability of the collected data can vary significantly depending on which scale is chosen and how the options (response choices) are constructed. Therefore, marketers must understand the characteristics of various response scales and use them appropriately.
The following are major response scale types frequently used in marketing surveys and their applications.
1. Likert Scale
- Key Features: Measures the degree of agreement/disagreement, satisfaction, importance, frequency, etc., regarding a specific statement. Typically, 5-point or 7-point scales are used. (e.g., 1=Strongly Disagree, 5=Strongly Agree)
- Advantages:
- Easy for respondents to understand and answer.
- Data collection and statistical analysis are relatively easy.
- Widely applicable to various topics.
- Disadvantages:
- Respondents may tend to avoid extreme choices and select middle values (central tendency bias).
- Social desirability bias (responding in a way perceived as socially desirable) may occur.
- The intervals between each option may not be perceived equally by respondents.
- Primarily Used Question Types & Specific Question Examples:
- Satisfaction Measurement: "Overall, how satisfied are you with the OOO product you recently purchased?"
(① Very Dissatisfied ② Dissatisfied ③ Neutral ④ Satisfied ⑤ Very Satisfied) - Agreement/Disagreement Measurement: "I believe our company's welfare system contributes to boosting employee morale."
(① Strongly Disagree ② Disagree ③ Neutral ④ Agree ⑤ Strongly Agree) - Importance Measurement: "How important do you consider the following factors when choosing a product? (Respond for each item)"
Price: (① Not at all Important ~ ⑤ Very Important)
Design: (① Not at all Important ~ ⑤ Very Important)
- Satisfaction Measurement: "Overall, how satisfied are you with the OOO product you recently purchased?"
- Points to Note When Constructing Options:
- Balanced Options: Must include positive, neutral, and negative responses in a balanced way. (e.g., 'Very Good, Good, Neutral, Bad, Very Bad' vs. 'Excellent, Very Good, Good')
- Clear Labels: The meaning of each option must be clear.
- Odd vs. Even Points: Odd-point scales offer a neutral option like 'Neutral,' while even-point scales do not allow neutrality, forcing a choice between positive or negative. Choose based on survey purpose.
- Option Order: Arrange in a logical order (e.g., negative → positive).
2. Semantic Differential Scale
- Key Features: Places a pair of opposing adjectives (or noun phrases) at opposite ends and asks respondents to rate their subjective feelings or images of a specific target between them. (e.g., Modern 1-2-3-4-5 Traditional)
- Advantages:
- Easy for respondents to express subtle differences in emotions or attitudes.
- Useful for multidimensional evaluation of brand image, product attributes, ad concepts, etc.
- Visually intuitive and can arouse respondent interest.
- Disadvantages:
- Selecting appropriate pairs of opposing adjectives is crucial and can be difficult.
- Interpretation of adjectives can vary depending on cultural background or individual experience.
- Analysis methods can be somewhat more complex than Likert scales.
- Primarily Used Question Types & Specific Question Examples:
- Brand Image Evaluation: "What do you think of the OOO brand?"
Trustworthy □ □ □ □ □ Untrustworthy
Innovative □ □ □ □ □ Conservative
Young □ □ □ □ □ Old - Product Attribute Evaluation: "Please evaluate the design of the newly launched smartphone."
Sophisticated 1--2--3--4--5 Clumsy
Light 1--2--3--4--5 Heavy
- Brand Image Evaluation: "What do you think of the OOO brand?"
- Points to Note When Constructing Options:
- Clear Antonym Pairs: Adjectives at opposite ends must have clearly opposite meanings.
- Number of Scale Points: Typically, 5-point or 7-point scales are used.
- Consistency or Alternation of Positive/Negative Direction: Consider placing positive adjectives on the left for all items, or alternating the positive-negative order for some items to reduce response bias.
- Whether to Specify Neutral Point: Decide whether to label the middle point as 'Neutral' or 'Average.'
3. Ranking Scale
- Key Features: Presents several items (items, attributes, brands, etc.) and asks respondents to rank them directly according to a specific criterion (preference, importance, urgency, etc.).
- Advantages:
- Clearly identifies the relative importance or preference among items.
- Effective for deriving priorities as respondents directly compare/evaluate each item.
- Disadvantages:
- If the number of items to rank is too large, it can increase the respondent's cognitive load and reduce response quality.
- The actual gap between ranks (e.g., the difference in preference between 1st and 2nd rank) cannot be known.
- Analysis can be somewhat complex (average rank, most frequent rank, etc.).
- Primarily Used Question Types & Specific Question Examples:
- Product Feature Preference: "Please select your top 3 preferred smartphone features in order. (1st, 2nd, 3rd rank)"
(Options: Camera Performance, Battery Life, Display Quality, Processing Speed, Storage Space) - Purchase Decision Factor Importance: "Please list the importance of the following factors when purchasing a car in order. (Most important first)"
(Options: Price, Fuel Efficiency, Design, Safety, Brand)
- Product Feature Preference: "Please select your top 3 preferred smartphone features in order. (1st, 2nd, 3rd rank)"
- Points to Note When Constructing Options:
- Limit Number of Items: Limiting the items to rank to preferably 5-7 helps reduce respondent burden.
- Clear Instructions: Clearly guide how to rank (e.g., from most important, from most preferred) and how many items to rank.
- Whether to Include 'Other' Item: Consider including an 'Other (specify)' item for the possibility of other important items not listed.
4. Nominal Scale
- Key Features: Used to simply classify or categorize responses. There is no order or magnitude meaning between categories. (e.g., gender, occupation, region of residence)
- Advantages:
- Easy for data classification and group division.
- Very easy to respond.
- Disadvantages:
- Numerical analysis (mean, standard deviation, etc.) is impossible.
- Only frequency or percentage analysis is possible.
- Primarily Used Question Types & Specific Question Examples:
- Demographic Information: "What is your gender?"<br>(① Male ② Female ③ Other)
- Product Usage Status: "Are you currently using the OOO product?"<br>(① Yes ② No)
- Points to Note When Constructing Options:
- Mutual Exclusivity: Each category should not overlap with others.
- Comprehensiveness: Should include all possible responses. If necessary, include 'Other' or 'Prefer not to answer' items.
5. Ratio Scale
- Key Features: Has an absolute '0' point, and ratio calculations between values are meaningful. All arithmetic operations are possible. (e.g., age, income, purchase frequency, usage time)
- Advantages:
- Contains the most information and allows for various statistical analyses.
- Precise measurement is possible.
- Disadvantages:
- For sensitive information (e.g., income), response refusal may be high.
- f asked as an open-ended question, the response range can be very wide, making data processing difficult.
- Primarily Used Question Types & Specific Question Examples:
- Usage Measurement: "How many hours in total did you use the OOO service last week? (Enter numbers only)"
- Purchase Amount: "How much did you spend on average per purchase recently? (Unit: KRW)"
- Points to Note When Constructing Options:
- Specify Units: Clearly present units such as amount, time, etc.
- Set Response Range (for multiple-choice): If it's difficult to ask as an open-ended question or you want to limit the response range, you can present it in multiple-choice format by dividing it into appropriate intervals. (e.g., income brackets) In this case, setting the intervals is important.
Considerations When Choosing a Scale:
- Nature of the Concept Being Measured: What do you want to measure – satisfaction, attitude, frequency, image, etc.?
- Understanding Level of Target Respondents: Can respondents easily understand the meaning and usage of the scale?
- Desired Level of Data Precision: Is it enough to grasp general trends, or do you need to measure fine differences?
- Data Analysis Method: How do you plan to analyze the collected data statistically?
- Response Time and Burden: How much time and effort will it take for respondents to answer the scale?
Problems with Incorrect Scale Usage:
Incorrect scale design, such as overlapping response options, omitting some responses, or providing unbalanced options skewed in a particular direction, seriously undermines data reliability. For example, if a question "How satisfied were you with our service?" only offers positive options like "Very satisfied, Satisfied, Slightly satisfied," dissatisfied customers cannot properly express their opinions, leading to distorted results.
In conclusion, the selection of an appropriate response scale and careful option construction are essential conditions for obtaining meaningful and reliable data through surveys. Marketers must fully understand the pros and cons of each scale and have the ability to choose the optimal scale according to the specific purpose and situation of the survey.