Sampling and Response: The experience of conducting an online student survey. Donna Poade
Overview of Presentation: l Background to the study l Online surveys as a research tool l Methodological issues in online survey research l Sampling and response l Our study l Tips for conducting online research
Background to the Study l First national survey of attitudes towards quantitative research amongst sociology and politics students across Universities in England and Wales l Whole study conducted November – March 2005/6 l Exploratory phase conducted followed by online survey l Builds on previous study of sociology departments and teaching staff
Online Survey Research l What is online survey research? l Expansion in using online resources as a medium for data collection l Types of online surveys: Web based or web page surveys Combination of two elements Surveys are the most implemented online research methodology in social research (Hewson et al 2003)
Online Survey Research Web Page Open webSimple Closed web attachment Hidden web URL embedded Source: Bradley (1999)
Advantages and Disadvantages of Online Survey Research AdvantagesDisadvantages TimeTechnical issues Cost efficiencyExpertise Access to wider population Threats to data validity
Methodological Issues l Methodological issues cited in the literature l Sampling issues l Response rates, including non response and response effects l Ethical issues, including privacy, confidentiality and informed consent l Practical issues such as incentives l Security l Technology
Sampling issues l Sampling bias l Sampling frames l Internal/External l Sampling approaches l Random Digit Dialling?
Sampling issues l Some sample sources: l /Website directories l Customer records l Pop up surveys l Harvested addresses l Registration Forms l Hardware, software and user ability
Response Rates l Inconclusive or inconsistent studies on online response rates l What is an acceptable response rate? l Response rates to online surveys 10% or lower l Response rates appear to be lower than for other more traditional methods
Our Study l Hard to reach population l Why did we choose this method? l Random sample of 34 sociology units in English and Welsh universities l URL embedded administered via a gatekeeper between January and April 2006
Whats the issue with non- response? l The Over-Surveying Effect l The sociology of non-response: (Tourangeau, 2006) l Non-contact (face to face) l Non-cooperation (too busy, self absorbed and erecting barriers) l Inability to provide data (physically or mentally) l Unfavourable societal developments on all three fronts (lack of civic engagement, obligations and more intrusions)
Response Rates in our Study l 738 responses equivalent to 12.8% for online survey l Prompted a boost by self completion paper surveys in some low responding units l Lowest reported online 1% (Basi 1999) l Response spikes attributed to the level and type of intervention l Less Socially desirable responses received
Tips for conducting online survey research l Consider appropriate sampling method l Use technical safeguards l Multi-modal approach l Response effects
Tips for conducting online survey research l Response effects: l Design (sponsor, personalised) l Incentives, l Pre-notification l Salience of topic l Piloting l Quality and repeat contact with respondents (including follow ups)
Summary l Methodological issues: the sample, the gate- keepers and the method l Never know the true sample bias l Are we right to be concerned? l What does the future hold for online survey research? l What is the best way of researching student populations?