"Given that Sauti za Wananchi is a longitudinal mobile phone survey, the number of survey instruments required in Sauti za Wananchi was larger than what is used in a traditional survey design. Uwazi at Twaweza designed the survey instruments and shared them with Ipsos Synovate (a research firm) for review. In several meetings during the initial weeks of the project, the understanding of the methodology was sharpened and resulted in a list of instruments required for the fieldwork" - Sauti za Wananchi: Collecting national data using mobile phones (SwZ Approach Paper) Issued in February 2013 by Twaweza.
"The survey was conducted by mobile phone. In order to ensure that the sample is representative of the whole of mainland Tanzania, mobiles phones and solar chargers have been distributed to participating households. Only households in areas lacking mobile phone network coverage are not represented. (Further details of the methodology are provided at the end of this brief, and at www.twaweza.org/sauti). The findings are based on the first call round conducted between 19 August and7 September 2015, with 1,848 respondents. At this point, the election campaigns were just beginning, with 8-10 weeks of campaigning still ahead. Respondents were selected using recognised methods of probability sampling as used by research institutions all over the world. The sample covers all regions of mainland Tanzania" - Let the people speak: Citizens’ views on political leadership (Brief No. 27) Issued in September 2015 by Twaweza.
"The target population for this survey was Tanzanians aged 18 years and above, of whom 1,836 living in urban and rural areas in the Tanzania mainland were interviewed. The margin-of-error attributed to sampling and other random effects is +/- 2.3 with a 95% confidence level. For demographic details, see the accompanying PowerPoint document.) Fieldwork was conducted between 5th and 22nd September 2015. Data was collected through face-to-face interviews using hand-held devices (smart phones)" - Press Statement (English)- Ipsos Tanzania Sept 2015.pdf
Dates of fieldwork: 05th - 22nd September 2015
Sample size: 1,836 (TanzaniaMainland Only)
Sampling Methodology: Random, Multi-stage stratified using PPS (proportionate to population size)
Population Universe: Tanzanian adults, aged 18 and above living in Urban and Rural areas (All 25 regions covered)
Data Collection Methodology: Face-to-Face interviews at the household level
Sampling Error: +/-2.3% with a 95% confidence level
(Note: Higher error-margins for sub-samples)
Interview Languages: English, Swahili
Ipsos employs quality control measures to ensure the highest possible integrity of obtained results/data
• For at least 20% of the interviews, a supervisor is present throughout
• Field manager visit at least 15% of the respondents in the sample at their households to confirm the interviews (i.e. back-checking).
• After the interview data are electronically transmitted to the Ipsos Server: an independent team then makes random phone calls to 40% of the respondents to confirm that the interviews were conducted with the said respondents (i.e.,telephonic back-checks).
• Mobile Data Collection Platform captures GPS location (latitudes and longitudes) of interviews conducted to verify the locations of the interviewers in the field.
• Logical data checks are made on selected questionnaire entries to ensure conformity to the sample’s statistical parameters.
- IPSO's 'TZ ELECTION POLL REPORT' (September 2015)