OSI PIPES STUDY Slovakia Data Collection
Overview - Slovakia 30 administrators 320 schools covered in total 513 families 1004 questionnaires collected –513 parents –454 teachers –37 headmasters
Finding administrators 30 administrators contracted - mostly teachers, but also parents, school counsellors, and university teachers received training (both oral and written materials) beforehand could contact the project team to clarify their questions/concerns about the Qs training: special attention was paid to possibly problematic/controversial points in Qs and the explanation of terminology
More about sampling administrators were assigned a geographical area to cover in the area, they selected possible respondents to cover both primary and secondary schools and rural and urban settlements could only rely on personal contacts to find respondents (no access to official teacher/student registers due to legislative restrictions)
Administering questionnaires face-to-face survey for all 3 types of Qs 1 questionnaire took about 1 hour to complete administrators read questions aloud and recorded responses, explained where necessary and helped with questions requiring the calculations of different amounts on a monthly/yearly basis
Confidential nature Before starting each face-to-face survey, administrators informed respondents that: –no personally identifiable data are collected –this is an international and scientific survey, not local and involving local policy and stakeholders –the identity of the school/family that the Q. is related to will only be used for internal purposes –after data entry all questionnaires will be returned to the local project manager and discarded
Challenges/Problems Qs were not received well by many respondents – too long/complicated (different figures on a monthly/yearly basis) some respondents (especially teachers and headmasters) were offended by Qs – felt as if suspected of illegal practices –mainly questions QS 15 K, QT 23, QT 18 a QT 39 –sometimes refused to answer all questions parents more willing to cooperate and give information
Challenges/Problems – cont. some administrators overestimated their abilities –promised to do more questionnaires than they actually returned completed as they found it very difficult to find enough respondents –the promised to delivered Qs ratio only 50% in the first round subsequently more administrators had to be found – delay in the process
Challenges/Problems – cont. the time for Qs administration – originally 6 weeks – extended by one month still slightly below the target numbers after extension, but enough to have a representative sample – checked with sociologist –planned 1240 in total 600 parents 600 teachers 40 headmasters
Representativity of questionnaires 513 parents –Population = families ( pupils of all the primary and secondary schools incl. special needs schools combined with 1.22 children/woman in Slovakia during period) –Confidence level 99%, confidence interval teachers –Population = internal teachers at all the primary, sec. and special needs schools (56413 at public schools only) –Confidence level 99%, confidence interval 6.03 (6.03) 37 headmasters –Population = 3181 public schools (primary, secondary, special needs) –Confidence level 99% (95%), conf. interval (16.02)
Threats to the reliability/validity of the measures difficulties with some personal questions (family income of parents, teachers and principals, giving the gifts, etc.) some figures may be unreliable –too difficult to recall all payments and indicate them on a monthly/yearly basis –we eliminated part of unreliable figures during data entry/cleaning – removal of evidently misinterpreted figures (in the multiple-item questions)
Parents – region/type of school DP2 – Region ValueCountPercent BA TT TN NR ZI BB PO KE Total QP23 – primary / secondary school ValueCount NA34 primary1309 secondary2170 Total513 ValueCountPercent NA10.19 urban rural Total
Teachers Regions ValueCountPercent BA TT TN NR ZI BB PO KE Total DT2 – Primary vs. secondary school –ValueCountPercent primary secondary Total DT3 - Urban vs. rural ValueCountPercent urban rural Total DT4 – Male vs. female respondent ValueCountPercent NA20.44 male female Total