Survey questionnaire designer with built-in bias detection

# Survey Instrument: Digital Literacy Self-Efficacy Among Adults 55+ ## Instrument Overview - **Construct:** Digital literacy self-efficacy (confidence in ability to use digital technologies effectively) - **Target:** Adults aged 55 and older - **Theoretical basis:** Bandura's Self-Efficacy Theory + Digital Literacy Framework (Hague & Payton, 2010) - **Estimated completion time:** 12-15 minutes - **Sub-dimensions:** (1) Basic Device Operation, (2) Information Navigation, (3) Digital Communication, (4) Online Safety, (5) Digital Problem-Solving --- ## Survey Items ### Section A: Basic Device Operation **Q1.** How confident are you in your ability to turn on a computer, tablet, or smartphone and navigate to an application you want to use? - *Type:* 5-point Likert (Not at all confident → Extremely confident) - *Bias check:* ✅ Clean — single behavior, neutral wording - *Sub-dimension:* Basic Device Operation **Q2.** I often need someone else to help me with basic tasks on my phone or computer. - *Type:* 5-point Likert (Strongly disagree → Strongly agree) - *Bias check:* ⚠️ **Social desirability risk** — admitting need for help may feel stigmatizing for older adults. Mitigate by framing the section with "People have different levels of experience with technology, and there are no right or wrong answers." - *Note:* **Reverse-coded item** (R) - *Sub-dimension:* Basic Device Operation **Q3.** When you need to install a new app or software update, how do you typically handle it? - *Type:* Multiple choice (I do it myself easily / I do it myself but it takes effort / I ask someone to help / I avoid updates entirely) - *Bias check:* ✅ Clean — behavioral, non-judgmental options - *Sub-dimension:* Basic Device Operation ### Section B: Information Navigation **Q4.** How confident are you in finding reliable health information online (for example, checking medication interactions or understanding a diagnosis)? - *Type:* 5-point Likert - *Bias check:* ⚠️ **Double-barreled risk** — "finding" and "reliable" are two separate skills. However, for self-efficacy measurement, the combined task reflects real-world behavior. Acceptable for this construct. - *Sub-dimension:* Information Navigation **Q5.** I can usually tell the difference between a trustworthy website and one that might be misleading. - *Type:* 5-point Likert (Strongly disagree → Strongly agree) - *Bias check:* ⚠️ **Social desirability risk** — people overestimate their ability to detect misinformation. Consider pairing with a behavioral task in pilot. - *Sub-dimension:* Information Navigation **Q6.** When you search for something online, how many of the following can you do? (Check all that apply) - [ ] Use quotation marks to search for an exact phrase - [ ] Filter results by date - [ ] Open results in a new tab - [ ] Evaluate whether a source is credible - [ ] Bookmark a page for later - *Type:* Checklist (behavioral inventory) - *Bias check:* ✅ Clean — concrete, observable behaviors **Q7.** 🔶 **ATTENTION CHECK** — For this question, please select "Somewhat confident" to show you are reading carefully. - *Type:* 5-point Likert (embedded as "How confident are you in using a device you have never heard of called a Zartonix Pro?") - *Bias check:* N/A — validity item. Fictional device name screens for acquiescence bias. ### Section C: Digital Communication **Q8.** How confident are you in sending an email with an attachment (such as a photo or document)? - *Type:* 5-point Likert - *Bias check:* ✅ Clean **Q9.** Using video calling (such as Zoom or FaceTime) is something I find easy to do. - *Type:* 5-point Likert (Strongly disagree → Strongly agree) - *Bias check:* ⚠️ **Leading — "easy"** implies it should be easy. Revise to: "I feel confident using video calling platforms (such as Zoom or FaceTime)." - *Sub-dimension:* Digital Communication **Q10.** I avoid sending messages or posting online because I worry about making mistakes. - *Type:* 5-point Likert - *Bias check:* ✅ Clean — captures avoidance behavior - *Note:* **Reverse-coded item** (R) **Q11.** Rate your comfort level with each of the following: [Matrix] - Sending a text message - Posting on social media - Joining an online group or forum - Making a video call - *Type:* Semantic differential (Very uncomfortable ←→ Very comfortable) - *Bias check:* ⚠️ **Acquiescence risk** in matrix format — respondents may select the same column throughout. Randomize row order. ### Section D: Online Safety **Q12.** How confident are you that you could recognize a phishing email (a fake email trying to steal your personal information)? - *Type:* 5-point Likert - *Bias check:* ✅ Clean — parenthetical defines term without leading **Q13.** I know how to create a strong password and use it to protect my online accounts. - *Type:* 5-point Likert - *Bias check:* ⚠️ **Double-barreled** — creating and using are two actions. Split into two items for the final version if factor analysis shows poor loading. **Q14.** Which of these have you done in the past 6 months? (Check all that apply) - [ ] Changed a password on an online account - [ ] Checked privacy settings on a social media account - [ ] Used two-factor authentication - [ ] Reviewed app permissions on your phone - [ ] None of the above - *Type:* Checklist (behavioral) - *Bias check:* ✅ Clean — behavioral, time-bounded, includes "none" option **Q15.** Technology companies do enough to protect people's personal information online. - *Type:* 5-point Likert - *Bias check:* ⚠️ **Leading — "enough"** carries implicit judgment. This item measures attitudes rather than self-efficacy. Consider removing or moving to a separate attitudes section. ### Section E: Digital Problem-Solving **Q16.** When something goes wrong with your device (for example, it freezes or an app crashes), how confident are you in troubleshooting the problem yourself? - *Type:* 5-point Likert - *Bias check:* ✅ Clean — example clarifies without leading **Q17.** If I can't figure out how to do something on my computer or phone, I usually give up. - *Type:* 5-point Likert - *Bias check:* ✅ Clean — captures persistence/self-efficacy - *Note:* **Reverse-coded item** (R) **Q18.** Rank the following from most to least likely when you encounter a technology problem: 1. Search online for a solution 2. Ask a family member or friend 3. Call customer support 4. Try different things until something works 5. Wait and hope it fixes itself - *Type:* Ranking - *Bias check:* ⚠️ **Order effect** — first option may be selected more often. Randomize presentation order. **Q19.** 🔶 **ATTENTION CHECK** — Please describe in one sentence what this survey is about. - *Type:* Open-ended (brief) - *Bias check:* N/A — comprehension and engagement check **Q20.** I feel comfortable learning new technology tools on my own, without formal instruction. - *Type:* 5-point Likert - *Bias check:* ✅ Clean — core self-efficacy item ### Section F: Open-Ended Reflection **Q21.** What is the biggest challenge you face when using digital technology? (2-3 sentences) - *Type:* Open-ended - *Bias check:* ⚠️ **Negative framing** — asking about "challenges" primes negative responses. Add a follow-up: Q22. **Q22.** What digital skill are you most proud of learning? (2-3 sentences) - *Type:* Open-ended - *Bias check:* ✅ Balances negative framing of Q21 --- ## Recommended Order & Rationale 1. **Warm-up:** Section A (Basic Device Operation) — easiest, builds confidence 2. **Core:** Sections B-D in randomized order across respondents to counterbalance fatigue effects 3. **Complex:** Section E (Problem-Solving) — most cognitively demanding 4. **Cool-down:** Section F (Open-Ended) — reflective, low pressure 5. **Demographics** last — age, education, years using technology, primary device, internet access type --- ## Cultural Sensitivity Notes for 55+ Population - Avoid jargon without definitions (always define terms like "phishing," "two-factor authentication") - Do not assume smartphone ownership — include tablet and desktop options - Avoid ageist framing ("for your age" or "despite being older") - Offer paper/phone alternatives for those with low digital access (ironic for a digital literacy survey — but necessary for representative sampling) - Test with participants from diverse socioeconomic backgrounds — digital access correlates strongly with income - Consider large font (14pt+) and high-contrast formatting for accessibility --- ## Pilot Testing Protocol 1. **Cognitive interviews** (n=8-10): Think-aloud protocol with diverse 55+ adults. Focus on item comprehension, response process, and emotional reactions. 2. **Test-retest reliability** (n=30, 2-week interval): Assess temporal stability of scores. 3. **Exploratory factor analysis** (n=150-200): Confirm the 5 sub-dimension structure. Remove items with cross-loadings >0.35 or communalities <0.40. 4. **Internal consistency:** Target Cronbach's α ≥ 0.80 per sub-dimension. 5. **Known-groups validity:** Compare scores between adults who have vs. have not completed digital literacy training. 6. **Readability check:** Flesch-Kincaid Grade Level ≤ 8 for all items.