African Journal of Science and Research,2016,(5)2:35-37 ISSN: 2306-5877 Available Online: http://ajsr.rstpublishers.com/ GIVING TWO ANSWERS TO THE SAME QUESTION: APPLYING SURVEY RE-INTERVIEWING TECHNIQUES IN QATAR. Elmogiera Elawad* A Diop, Mohamed Agied, Isam Abdelhameed. Social and Economic Survey Research Institute- SESRI,Qatar University ,Qatar. Email: Elmogiera@qu.edu.qa Received:17,Feb,2016 Accepted: 27,Mar,2016 Abstract In order to evaluate the quality of its survey data, SESRI implemented since 2010 a system of re-interviewing a sub-sample of respondents. These re-interviews usually take place not more than six days from the date of the first interview for the original survey production. Results from these interviews are used as an instrument to analyze non-sampling error and to estimate response inconsistencies, to identify reasons for discrepancies for selected questions for the same respondents, as well as to determine interviewer cheating. Applying the Gross Difference Rate (GDR), Index of Inconsistency (IOI) and Reliability Rate (RR) to estimate variances in respondents’ answers, this paper presents the results of the re-interviewing study of the 2012 Qataris’ Attitudes Towards the Foreign Workers in Qatar. The results are inte nded for researchers and data collection organizations to improve the quality of data they collect and use. Keywords:Survey,Quality, Index of Inconsistency and Reliability Rate. INTRODUCTION Since 1984 census in Qatar state; a lot of surveys in various aspects and for different purposes have been conducted, however there is few information about methodological studies in survey data quality, in particular variances in respondent’s answers, interviewer falsification or interviewing errors’. In general response errors often occur as a result of problems related to questionnaire design, interviewer, respondent or processing errors, and there are many methods for measuring errors resulting from the respondent and interviewers for non-sample errors in survey research, which have a large impact on data quality, and there are several techniques to supply these methods such as check studies, change the data collection methodology, change the questions or the questionnaire as a whole for a selected sub sample from the original sample or use Cognitive Research Laboratories, One of the most important methods to estimate respondent and interviewers errors is re-interview, (Daniel Kasprzyk,1999), which have been usually used to measure data quality of surveys (Forsman and Shriner 1991). Participation in polls in Qatar is similar to all studies in the world which is voluntary and gurantee the confidentiality of the data, but there is no incentives for participation, so all participants in the Qatari Attitude towards foreign worker (QAF) production (CAPI) and re- interview (CATI) were willingly involved. STUDY AREA The objectives of this paper is to present the findings of the re- interview study conducted in Qatari Attitude towards foreign worker survey (2012) in Qatar state , to collect information on :  Response Inconsistencies  Identify reasons for discrepancies for identified questions  Determine interviewer cheating, and.  Estimate re-interview response rate using CATI (Computer Assistance Telephone Interview) METHODOLOGY In a period not more than six days from the date of the first interview for the original survey production QAFS 2012 (Qatari attitude towards foreign worker survey 2012), and not less than two days, callback was executed to a selected random subsample from the respondents in the survey, all selected respondents in the callback were asked to answer Four questions from the original questionnaire . Accordingly; the surveying mode and interviewers have been changed in the re-interview. The similarity between the original production sample and the re-interview subsample has been taken into account, not only the population distribution according to zones and areas; but also the completed interviews for each (interviewers to verify the visit). The QAF survey was conducted in the period from December 1st to December 27, 2012 and 2394 Qatari Nationals were interviewed using computer assistance person interview CAPI, and the re-interview was conducted using computer assistance telephone interview (CATI) from December 8 to December 25, 2012 for 387 respondents which is about 15% of the total sample. All respondents were informed about the purpose of the study and were not reminded about their previous answers. Analytic methodology The re-interview method is an influential instrument in non- sampling error analysis. The Gross Difference Rate (GDR), Index of Inconsistency (IOI) and Reliability Rate (RR) are the indicators we used in the re-interview to estimate variances in respondent answers The gross difference rate (GDR), is the average squared difference between the original interview and re interview responses (see Biemer and Forsman, 1992 ; Biemer 2004) . The GDR divided by 2 is an unbiased estimate of simple response variance (SRV), we estimate the GDR using the following formula: 2 2 2 1 M M GDR   And we also calculated the Index of inconsistency (Brick, Rizzo and Wernimont (1997)) , Which measures the proportion of the total population variance attributed to the simple response variance. Hence, 2 2 1 2 S S GDR IOI   where 1 2 S is the sample variance for the original interview and