As an organization, we spare no effort in conforming to international standards to ensure client satisfaction. We follow rigorous survey methods to draw samples representative of the political and social demography of the universe that is polled.

Sample Design

In multi-stage systematic stratified random sampling method, respondents are sampled from latest electoral rolls or any other sample universe and the surveys are executed through multi- stage sampling techniques. Selection of sampling techniques are made on the basis of their controls on the sample distribution they cover geographically, making sure it is spread over the spatial dimension similar to the target population.

With Systematic Random Sampling, we take a list of every member of the sample frame. From the list, we randomly select the first sample element i.e. our first respondent number from the list. Thereafter, we select every nth element on the list with a constant application of the interval.

With cluster sampling, the sample frame is divided into separate groups or clusters. In electoral surveys, every polling station that has the same polling booth is assigned to one, and only one group or cluster. A sample of clusters is chosen, using a probability method.

Questionnaire Design

The questionnaire is designed as a Pre-coded Standard Structured Interview Schedule. By and large, we mainly focus on four points while designing a questionnaire: content, phrasing, format used to collect responses with the appropriate use of logic flows, and use of precise, lucid language. There is optimal use of technology through Search and Navigation buttons. For electoral surveys, a dummy EVM is used for a mock voting drill, wherever possible using the names of nominated candidates against the party symbols to authenticate the experience.

Care is taken in designing the questions so that each respondent receives the same stimuli. Standard definitions and explanations for each question is decided after thorough discussions keeping in consideration the objectives of the survey. This also helps field investigators in handling clarifications about questions.

Face to face interviews conducted
by field investigators

Extensive training sessions are conducted for field investigators through workshops and mock sessions that include face-to-face interaction in administering the questionnaire and handling different kinds of respondents. After the training, data is collected through CAPI (Computer Assisted Personal Interviews) with GPS-enabled tablets that facilitate logic checks, skip patterns and validations during the interview. GPS coordinates of investigators and geo-tagging of respondents, that record each interview’s start time, end time and geographic location, work as a reliable tool to monitor the survey process, and go a long way in ensuring accuracy as well as precision of data collection.

The authenticity of the survey findings is bolstered by live-time data updates. The in-house checking team at Cicero ensures that places where substitution of respondents has taken place, a strict criterion of adhering to a similar geographic, social and demographic profile, has been followed..

Checking and cleaning of data

Data is cleaned to remove interviews of respondents who do not meet the target criteria, who are speeders (time taken to complete a questionnaire that is tracked through GPS of the investigator), and set responses.

Processing and analyzing data

Our dedicated team of experts with vast experience drawn from various fields help us to process and analyze the data using SPSS (Statistical Software for Social Sciences). The findings are presented in tabular as well as graphical forms. The results are documented in a report, annotated by descriptive statistics and commentary, which is submitted to the client.

Click on a step to find more about it.

Apart from sample survey methods, Cicero also employs Observational approach as a Descriptive Study that helps to complement the findings of the survey. We are also one of the earliest organizations in India to harness the potential of Controlled Survey Experiments through Randomized Control Trials (RCTs). RCTs are quantitative, comparative and controlled survey studies in which investigators study two or more interventions in a series of individual respondents. Through Focused Group Discussions, we arrive at meanings that add to the survey findings by focusing on the ‘why’ and ‘how’ aspects of an issue. In-depth information is gleaned through interviews of Key Informants that include stakeholders at different levels. Our expertise in a plethora of methods enables us to customize our approach depending on the requirements of the project and the specifications of the client. This, in turn, ensures high quality of research and greater accuracy in findings.