B= â.457***. â2,7% p. mail. +17,0% per prev. part. Demographic Margins. 0. 10,000. 20,000. SoSci Panelists. Education. Employed. Freelancer. Not in Job. 49%.
soSci
The Noncommercial Online Access Panel
A Super Convenience Sample
Solid Growth since 2009 – thanks to Thousands of Surveys
Source: Internal statistics SoSci Survey/Panel 01/2011-12/2011 1 Measured during 10/2011-01/2012
Dec Jun Dec 2009 2008
Female Online Research
Survey Load Limits
Especially online convenience samples often report a surplus of women. Typically 55% to 70% of the participants are female e.g., 8,11,13. In the SoSci Panel 58% of the panelists disclosing their gender are women (students: 63% female). Those who actually click survey links are 60% female n/a (n=14,363 in 09/2011-02/2012), with significant deviations between different studies and questionnaire topics.
Demographic Margins 0
10,000
48%
Commercial panels‘ participants join under the promise of getting paid for completing questionnaires. Positive effects of uncontingent incentives on return rate in random samples are well known 2,5 and research on data quality 3,4,9,14,15 shows only limited detrimental effects 7,12. This, however, cannot compensate for the lack of internationally published research regarding the effects of paying people for being available for surveys. Topic interest is one of the best predictors for data quality 6 – and it is what motivates people to participate in the SoSci Panel and to accept invitations to specific surveys.
Not in Job
40%
Compared to a commercial panel, two differences are obvious: The SoSci Panel is free of charge and it does not provide any representativeness regarding demographic variables. But the most important difference may lie in participants‘ motivation.
Employed Freelancer
7%
5%
SoSci Panelists 49%
29%
54%
No traditional Access Panel
20,000
63%
Education
SoSci Panel Participants revealing gender/occupation (n=51,588, N=70,844, 02/2012)
SoSci Panelists 0 10,000
70+
B= 1.260*** (p0= 22%) +17,0% per prev. part.
invitations
B= −.457*** within last 12 months −2,7% p. mail
time since last mail minimum 2 weeks
participation
probability, pM= 27,8%
B= .000 n.s.
Do regular survey invitation exhaust a panel’s participants? The SoSci Panel limits survey load to 4 invitations per year and a minimum lag of 2 weeks between two invitations. Logistic regression of participation (10/2011-02/2012, N=36,321 mails) shows a weak negative effect (ΔR²=5%) of invitations sent within the last year and no effect of the time lag when controlling for previous survey participations (R²1=9%, R²2=14%).
Finished College
60-69
Matriculation Standard
50-59
D
Secondary School
40-49 30-39
Primary School
20-29
Still in School
14-19 0 5,000 10,000 Sample (N)Onliner 10 German SoSci Panel Participants (N=42,416) revealing their age (n=42,323, 02/2012) revealing their formal education (n=39,479)
CH 4,000 8,000 12,000 0 Sample (N)Onliner 10
(N)Onliner Sample 10 representing the German Population (N=30,718) Onliners 10
Church, A. H. (1993). Estimating the effect of incentives on mail survey response rates: A meta-analysis. Public Opinion Quarterly, 57(1), 62–79. Davern, M., Rockwood, T. H., Sherrod, R., & Campbell, S. (2003). Prepaid monetary incentives and data quality in face-to-face interviews. Public Opinion Quarterly, 67(1), 139–147. 4 Godwin, R. K. (1979). The Consequences of Large Monetary Incentives in Mail Surveys of Elites. Public Opinion Quarterly, 43(3), 378. 5 Göritz, A. S. (2006). Incentives in Web Studies: Methodological Issues and a Review. International Journal of Internet Science, 1(1), 58–70. 6 Groves, R. M., Couper, M. P., Presser, S., Singer, E., Tourangeau, R., Acosta, G. P., & Nelson, L. (2006). Experiments in producing nonresponse bias. Public Opinion Quarterly, 70(5), 720–736. 7 Hansen, R. A. (1980). A Self-Perception Interpretation of the Effect of Monetary and Nonmonetary Incentives on Mail Survey Respondent Behavior. Journal of Marketing Research, 17(1), 77–83. 8 Papacharissi, Z., & Mendelson, A. (2011). Towards a new(er) sociability: uses, gratifications, and social capital on Facebook. In S. Papathanassopoulos (Ed.), Media perspectives for the 21st century (pp. 212–230). London: Routledge. 9 Petrolia, D. R., & Bhattacharjee, S. (2009). Revisiting incentive effects. Public Opinion Quarterly, 73(3), 537–550. 10 Peterhans, M., & Sagl, S. (2011). (N)Onliner Atlas 2011: Eine Topographie des digitalen Grabens durch Deutschland: Königsdruck. 11 Quan-Haase, A., & Young, A. L. (2010). Uses and Gratifications of Social Media: A Comparison of Facebook and Instant Messaging. Bulletin of Science, Technology & Society, 30(5), 350–361. 12 Schneider, K. C., & Johnson, J. C. (1994). Link between Response-inducing Strategies and Uninformed Response. Marketing Intelligence & Planning, 12(1), 29–36. 13 Sheldon, K. M., Abad, N., & Hinsch, C. (2011). A Two-Process View of Facebook Use and Relatedness Need-Satisfaction: Disconnection Drives Use, and Connection Rewards It. Journal of Personality & Social Psychology, 100(4), 766–775. 14 Singer, E., van Hoewyk, J., & Maher, M. P. (1998). Does the payment of incentives create expectation effects? Public Opinion Quarterly, 62(2), 152–164. 15 Trussell, N., & Lavrakas, P. I. (2004). The influence of incremental increases in token cash incentives on mail survey response. Public Opinion Quarterly, 68(3), 349–367. 3
since 07/2011, max. 3
Jun Dec 2011
SoSci Panelists 0 20,000 10,000
20,000
References 2
Jun Dec 2010
prev. participations
66,788
• Quasi-experimental settings (e.g., testing effects for interactions with age and gender) • Huge experimental settings
56,041
Providing extra heterogeneity and large availability, the SoSci Panel also offers some specific uses:
2.072 new Panelists p. m. (effective opt-in rate 5.7%1)
41,932
• Qualitative research and pilot studies • Research on general processes (e.g., perception) • Testing hypotheses in experimental settings
37% of the Survey Projects show the Panel Invitation 1 22,726
Lacking representativeness, the panel is useful in similar situations as convenience samples:
Source: Internal statistics SoSci Panel
88,640 Interviewees p. m.
18,483
The SoSci Panel’s Uses
Panelists registered in the SoSci Panel
12,586
The SoSci Panel recruits its participants from samples surveyed in uncounted online studies. From such a disperse source of panelists, a super convenience sample emerges that cannot only outnumber traditional student samples, but exceeds them in heterogeneity regarding age, geographic and professional background, interests, etc.
550 Surveys per month (p. m.)
32,133
Convenience samples play an indispensable role in social science: Essential chapters in the history of empirical research are based on surveying and observing students, their families, and friends. In the Internet age, users of social network sites joined the canon of easily available subjects. Such samples are beneficial whenever the study design does not require a sample representative for (typically) a country’s population.
A
Population: 12/2007 D, 07/2011 A, 12/2010 CH SoSci Panel Participants D/A/CH (N=51,318) disclosing postal code (n=40,528, 02/2012) Population (official statistics, scale 1:2400)
Researcher Dominik Leiner IfKW, LMU Munich +49(89)2180-9440
SoSci Panel Dominik Leiner SoSci Survey www.soscisurvey.de