Recruitment of a Probabilitybased Internet Panel in France. The ELIPSS Pilot Study

Transcription

1 Recruitment of a Probabilitybased Internet Panel in France. The ELIPSS Pilot Study Anne Cornilleau ESRA Conference, July

2 ELIPSS in few words Part of an infrastructure project focused on data for the social sciences Representative probability based internet panel Pilot study : individuals (5 000 in 2015) Monthly questionnaires designed by researchers (up to 30 minutes per month) Internet access given through tablets & 3G data plans to each panelist

3 Sample design Housing Unit (HU) sample drawn from rotating Census by the French National Statistical Institute (INSEE) Stratified two-stage probability sample HU including a HU reserve sample An individual randomly selected within the household Aged between 18 and 75 French-reading

4 Panel membership Formal agreement signed by each panelist Regular participation to surveys of the selected person Personal use of tablet in accordance with law 1026 members in June 2013 RECR = 29.3 % 23 inactive members

5 Recruitment design : main sample In house recruitment June 2012 August 2012 Private survey organisation Oct Feb HU FIRST Invitation Letter Reminder Letter(s) Follow ups by Phone 2217 Non contacts Refusals F2F Online form Postal (Reply Card) Phone 248 answers 421 answers 28 answers

6 Recruitment design : reserve sample Private survey organisation January April HU Advanced Letter Phone F2F

7 Recruitment outcomes Premilinary Results - please do not cite Main Sample Reserve Sample (N = 987) In house F2F reissue (N = 2667) Phone > F2F (N = 612) F2F only (N = 375) Total adresses (N = 4411) 1. HU agreement 15.2% 19.1% 28.4% 33.1% 30.1% 2. HU refusal 17.3% 38.9% 38.6% 27.7% 33.6% 3. Unknow eligibility 64.4% 24.7% 4.1% 10.1% 17.1% 4. HU ineligibles 3.0% 17.2% 28.9% 29.1% 19.2% HU RR1 1/(1+2+3) 15.7% 23.1% 40.0% 46.6% 37.2% #Panelists (IND Agreement/1) 74.7 % 76.9% 80.3% 85.5% 77.6%

8 Recruitment outcomes Premilinary Results - please do not cite Main Sample Reserve Sample (N = 987) In house F2F reissue (N = 2667) Phone > F2F (N = 612) F2F only (N = 375) Total adresses (N = 4411) 1. HU agreement 15.2% 19.1% 28.4% 33.1% 30.1% 2. HU refusal 17.3% 38.9% 38.6% 27.7% 33.6% 3. Unknow eligibility 64.4% 24.7% 4.1% 10.1% 17.1% 4. HU ineligibles 3.0% 17.2% 28.9% 29.1% 19.2% HU RR1 1/(1+2+3) 15.7% 23.1% 40.0% 46.6% 37.2% #Panelists (IND Agreement/1) 74.7 % 76.9% 80.3% 85.5% 77.6%

9 Recruitment outcomes Premilinary Results - please do not cite Main Sample Reserve Sample (N = 987) In house F2F reissue (N = 2667) Phone > F2F (N = 612) F2F only (N = 375) Total adresses (N = 4411) 1. HU agreement 30.0% 28.4% 33.1% 30.1% 2. HU refusal 33.4% 38.6% 27.7% 33.6% 3. Unknow eligibility 18.8% 4.1% 10.1% 17.1% 4. HU ineligibles 17.8% 28.9% 29.1% 19.2% HU RR1 1/(1+2+3) 36.5% 40.0% 46.6% 37.2% #Panelists (IND Agreement/1) 76.1% 80.3% 85.5% 77.6%

10 Recruitment outcomes Premilinary Results - please do not cite Main Sample Reserve Sample (N = 987) In house F2F reissue (N = 2667) Phone > F2F (N = 612) F2F only (N = 375) Total adresses (N = 4411) 1. HU agreement 30.0% 28.4% 33.1% 30.1% 2. HU refusal 33.4% 38.6% 27.7% 33.6% 3. Unknow eligibility 18.8% 4.1% 10.1% 17.1% 4. HU ineligibles 17.8% 28.9% 29.1% 19.2% HU RR1 1/(1+2+3) 36.5% 40.0% 46.6% 37.2% #Panelists (IND Agreement/1) 76.1% 80.3% 85.5% 77.6%

11 Recruitment outcomes Premilinary Results - please do not cite Main Sample Reserve Sample (N = 987) In house F2F reissue (N = 2667) Phone > F2F (N = 612) F2F only (N = 375) Total adresses (N = 4411) 1. HU agreement 30.0% 28.4% 33.1% 30.1% 2. HU refusal 33.4% 38.6% 27.7% 33.6% 3. Unknow eligibility 18.8% 4.1% 10.1% 17.1% 4. HU ineligibles 17.8% 28.9% 29.1% 19.2% HU RR1 1/(1+2+3) 36.5% 40.0% 46.6% 37.2% #Panelists (IND Agreement/1) 76.1% 80.3% 85.5% 77.6%

12 Comparison with national statistics Preliminary Results - please do note cite Age Groups Census (Jan. 2013) ELIPSS Panel Members % 8.8 % % 19.0 % % 24.6 % % 21.9 % % 17.0 % % 8.7 % Gender Census (Jan. 2013) ELIPSS Panel Members Women 51.0 % 51.3 % Men 49.0 % 48.7 % Internet access at home 2012 INSEE Survey ELIPSS Panel Members Yes 83.2 % 92.3 % No 16.8 % 7.7 %

13 Incentives experiment Preliminary Results - please do note cite 2000 HU of the main sample received a 10 voucher enclosed with the first invitation letter Positive impact on response: 37.9 % vs 32.2 % Positive impact on agreement: 18.8 % vs % Only for HU using the reply card HU Refusal HU Agreement Total No incentive (n = 147) 62.6 % 37.4 % 100 % Incentive (n = 250) 40.4 % 59.6 % 100 % Total (n = 397) 48.6 % 51.4 % 100 %

14 First Surveys Since December 2012 Tutorial to use the tablet; digital practices >> 91.6 % Elipss annual study Part 1 (socio demographic profile) >> 91.2% Part 2 (health, politics, social integration ) >> 86.9 % Cultural participation and preferences >> 86.9 % Planned studies Sexual and reproductive health Electoral choices and political values (longitudinal) Attitudes towards environment and Schwartz values Partnership arrangements and views on family issues

15 Participation in surveys and in ELIPSS Participation in survey before ELIPSS No, was never ask to (n = 302) 32.7 % No, always refused (n = 125) 13.5 % Yes (n = 497) 53.8 % Total (n = 924) Main reasons for participating in ELIPSS (2 answers) Free tablet (n = 571) 61.8 % Trust in ELIPSS institutions (n = 428) 46.3 % Originality of the project (n = 337) 36.5 % Interested in research (n = 297) 32.1 % Free internet access (n = 123) 13.3 %

16 Questionnaire design

17 Concluding remarks Recruiting directly by face-to-face more effective Cost & addresses quality issues Age and gender composition close to general population On-going work on socio-economic representativeness Giving tablet and 3G plan to everyone Incentive in 2012, but in 2014?

18 Questions? Project website: Panelist website:

19 Timeline In-house Invitation, recruitment reminders June July August Sept. Oct. Nov. Dec Jan. Feb. March April May Follow-up calls Advanced Letter Survey organisation recruitment Telephone company Contract negotiation Face-to-face re-issue (main sample) Phone training Reserve sample recruitment Surveys First Survey ELIPSS Annual Survey