Crafting survey questions that accurately measure motivation, anxiety, and satisfaction using Likert scales.
Start LearningIt is the process of creating individual survey questions (or statements) that accurately measure a particular construct.
In quantitative research, especially with Likert scales, well-crafted items are critical for:
Items are grouped into constructs (also called variables, scales, or dimensions) to measure broader traits.
Construct: Anxiety
"I feel tense before giving a public presentation."
The 5 pillars of effective survey questions
Use simple, unambiguous language.
“I enjoy doing research”
“I find research somewhat tolerable under certain conditions”
Each item must clearly relate to the construct.
“I feel confident using English in class” (Self-efficacy)
Avoid double-barreled items (asking two things).
“I enjoy reading and writing academic papers”
Correction: Split into two items.
Mix positively and negatively worded items.
“I feel motivated to learn”
“I often give up when learning is difficult”
Avoid emotionally charged or leading language.
“Research is a waste of time” (Too biased)
Including a mix prevents response bias (e.g., ticking “Agree” for everything).
A construct is measured by multiple items. Here is a sample survey.
I get nervous using unfamiliar technology.
I avoid digital tools whenever possible.
I feel anxious when I must use SPSS.
I feel confident troubleshooting tech problems.
Participants will skim or skip if they have to read a paragraph.
Ex 1: "I believe that despite the challenges of modern grammar instruction, it is ultimately beneficial for students..."
Ex 2: "When considering my future career path in linguistics, I often wonder if the theoretical frameworks..."
Words like "often" or "rarely" mean different things to different people.
Ex 1: "I read research papers often." (Daily? Weekly?)
Ex 2: "I rarely use English outside of class." (Once a month? Never?)
If they don't know the word, they can't answer truthfully.
Ex 1: "I feel my interlanguage fossilization is increasing."
Ex 2: "The construct validity of this test is high."
Asking two things at once forces an inaccurate answer.
Ex 1: "I enjoy studying grammar AND listening to music." (What if I only like one?)
Ex 2: "The instructor is knowledgeable AND approachable."
Double negatives confuse the brain.
Ex 1: "I do not believe that not studying is a good idea."
Simply confusing phrasing.
Ex 2: "It is not true that I dislike English."
Start with 10–15 items per construct. After pilot testing and reliability analysis (e.g., Cronbach's Alpha), you can reduce it to the most effective 3–5 items.