This preview shows page 1. Sign up to view the full content.
Unformatted text preview: SelfAdministered SelfAdministered Surveys
Mail and Internet Two Questions to Ask Yourself Two Questions to Ask Yourself
► Who are my respondents? ► Is my research question better answered by selecting one method over another? If the answer is NO, then consider TIME and MONEY! Selecting a Method Selecting a Method
► 1. Three broad factors to consider: Administration or resource factors 2. Questionnaire Issues 3. Data Quality Issues Factor 1: Mail Factor 1: Mail
Administration Cost Length of data collection Distribution of sample Mail Low Long (10 weeks) May be wide Factor 1: Internet Factor 1: Internet
Administration Cost Length of data collection Distribution of sample Internet Very low Very short (13 days) Wide Factor 2: Mail Factor 2: Mail
Questionnaire Issues Length Questionnaire Complexity Complexity of questions Openended questions HH/personal records Rapport Sensitive topics Nonthreatening quest. Mail Short – Med. (412 pgs) Simple Simple to moderate Poor Very good Fair Good Good Factor 2: Internet Factor 2: Internet
Questionnaire Issues Length Questionnaire Complexity Complexity of questions Openended questions HH/personal records Rapport Sensitive topics Nonthreatening quest. Internet Short (15 minutes) May be complex Simple to moderate Fair to good Very good Poor/Fair Poor/Fair Good Data Quality Factor 3: Mail Factor 3: Mail Mail Sampling frame bias Response rates Response bias Info on refusals and non contacts Control of response Quality of recorded responses Usually low Poor/Good Med./High Fair Poor Fairly good Factor 3: Internet Factor 3: Internet
Data Quality Sampling frame bias Response rates Response bias Info on refusals and non contacts Control of response Quality of recorded responses Internet Low to High Poor/Good Med./High Fair Poor Fair/Good Mail Surveys Mail Surveys
► Components may include: An advance letter (optional) Cover letter Questionnaire Return response envelope Incentive Special instructions/Questionnaire map/Examples Thank you/Please complete postcard Mail: Cover Letter Mail: Cover Letter
► Cover Letter should include: Purpose of the study Sponsor Statement of confidentiality Instructions about who is to complete questionnaire Time to take Incentive When to return Who to contact if there are questions Mail Questionnaire Mail Questionnaire
► Give your questionnaire a title ► Provide clear instructions for completion (simple) ► Construct questionnaire such that: Questions are simple to moderate construction Transitions between topics are recommended There are few openended questions Demographics are placed at the rear Skip instructions are clear Advantages of a Mail Survey Advantages of a Mail Survey
► Less expensive ► Able to collect record data ► The use of visual aids ► Sensitive topics may be covered ► Data collection period is fairly pedicatable Disadvantages of Mail Surveys Disadvantages of Mail Surveys
► Response bias (To compensate for response bias – get to know as much as possible about nonrespondents) Questionnaire Issues – Length has an impact R can look over quex prior to completing Little control over the order of administration Little control over who actually completes quex Topic saliency Complexity of questionnaire Openend are less likely to be complete Verifying that R understand instructions ► Internet Surveys Internet Surveys
► Components may include: An advance email/letter (optional) Participation invite with internet link or address Thank you/Please complete email Incentive Advance Email/Cover Letter Advance Email/Cover Letter
► Advance/Cover email should include: Purpose of the study Sponsor Statement of confidentiality Instructions for accessing website Time to take Incentive PIN – may be in advance email or embedded in URL Internet Surveys Internet Surveys
► First Screen Welcome Brief introduction may restate some of what was in advance email (purpose of study, time it takes to complete, sponsor ► Multiple reminders necessary Questionnaire construction must take into account R’s computer literacy, education level, hardware and software use Internet Questionnaire Internet Questionnaire ► Questionnaire may include: Complex skip patterns Visual aids, such as, popup instructions, help screens, dropdown lists, pictures, video and audio Detailed responses to openends Short (1012 minutes) Must be selfexplanatory ► Social Interface Theory: Internet questionnaires may include humanizing cues that allow the R to interface with the computer like they would with an interviewer Advantages of an Internet Survey Advantages of an Internet Survey
► Very low cost ► Speed of data collection Day 1 will yield up to 30% of your responses Day 2, 50% 10 to 20 days ► Allows large sample sizes: Both national and international samples Disadvantages of an Internet Survey Disadvantages of an Internet Survey
► Computer/internet access 57%/50% respectively ► Not able to design a general pop. prob. Survey ► Lack of a good sample frame ► Low response rate ► Response bias Internet Disadvantages cont. Internet Disadvantages cont. ► Internet questionnaire – May not appear the same to all Rs due to hard/software ► In situations where the questionnaire is constructed such that the R must answer questions before moving on there may be an increase number of DK or RF ...
View Full Document
This note was uploaded on 06/11/2010 for the course SOC 481 taught by Professor Burris during the Summer '09 term at University of Illinois at Urbana–Champaign.
- Summer '09