Online Survey Tools and their use in Public Health
Vicky Ng, PhD
Centre for Public Health and Zoonoses (CPHAZ), University of Guelph
April 5, 2012
Presentation Outline
• Online trends in Canada
• Advantages and disadvantages
• Online survey tools to overcome biases
• Recruiting for an online survey
• Software/IT requirements
• Online survey use in public health
Internet access in Canada
Source: World Bank, World Development Indicators (accessed April 2, 2012)
Internet usage in Canada
Source: Globe and Mail (published online March 8, 2012)
Survey methods in Canada
Source: Report for MRIA Gold Seal Research Agency Member Firms (Member and Non-member Results, 2007 to 2010).
Advantages to Online Surveys (1)
• Increase in internet access/usage
• Internet demographics rapidly changing; usage near universal (age, gender, income, region)
• Automatic data collection
• Real-time accumulation of data (quota/analysis)
• Time- and cost-efficient (paperless)
• Errors can be corrected
• Respondent errors are flagged automatically (no missing or out-of-range responses)
Advantages to Online Surveys (2)
• Higher response rate (no time/geographic constraint, larger sample size)
• Secure and anonymous (honesty, REB)
• Access to hard-to-reach individuals/groups
• Eliminates interviewer bias or error
• Complex functions (auto-skip, personalization of questions, quota control, multiple survey versions, choice randomization)
Disadvantages to Online Surveys
• No access to those without internet
• Target population may not be tech-savvy
• Bias towards those who tend to respond to online surveys (young to middle-age, white females)
• No interviewer to clarify and probe, possibly leading to less reliable data
• Panel integrity (inconsistent responders, speedy responders, repeat responders)
• Technical issues (IT set-up, server problems, program bugs)
Online functions to overcome biases
• Quota control (representative sample, eligibility)• Auto-skip and personalized questions/graphics
(streamline process = higher integrity responses)• Multiple survey versions/choice randomization
(ordering bias)• Consistency check questions (disqualify inconsistent
responders)• Time stamp (disqualify speedy responders)• IP addresses/unique IDs (disqualify repeat
responders)• Purge “straight line” responders
Recruiting for an online survey
1. Web link in print publications/pamphlet (passive)
2. Web link in e-pubs or websites (semi-active)
3. Direct email (active)
• Mailing list (targeted organizations)
• Online panel providers (opt-in panels)
4. Direct contact/web (active+)
• Conferences, hospitals, public outreach programs
• Online panel providers (custom panels)
Online Panel Providers (OPP)
• Opt-in panels:– Pre-recruited/pre-screened respondents who generally
complete a number of surveys over time– Incentive-based– Large and broadly-based (country-specific)– Specialist panels available (auto, mobile, healthcare)
• Custom panels:– Randomly recruited from the general population– Hybrid internet-telephone sampling– Refreshed constantly, cannot opt-in (randomly selected)
Software/IT requirements
• No software necessary if you outsource recruitment, survey design, hosting and data collection entirely to OPP
• Simple software if you are only outsourcing a portion of the study (e.g. FluidSurveys, LimeSurvey)
• Sophisticated software for specialized methods (e.g. Sawtooth Software, but check with IT!)
• Internet access• Server to host the survey and data• Computer to download and analyze data• Email invites (spam, unsubscribe, generic email)
Examples of Public Health Surveys
• Online surveys are versatile and most studies can be conducted online:
– Assess health and educational services
– Explore the provision of health services
– Internal evaluation by public health officials
– Identify public health problems and gaps in information
– Population health study (e.g. burden of illness, Ontario Health Study)
– Public education/feedback