- Tytuł:
-
Analizy weryfikacyjne – przeszłe i obecne doświadczenia badawcze
Verification analysis in social surveys. Past and present research experiences - Autorzy:
-
Grzeszkiewicz-Radulska, Katarzyna
Krzewińska, Aneta - Powiązania:
- https://bibliotekanauki.pl/books/28836008.pdf
- Data publikacji:
- 2017
- Wydawca:
- Uniwersytet Łódzki. Wydawnictwo Uniwersytetu Łódzkiego
- Opis:
-
Prezentowana w tym opracowaniu refleksja wpisuje się w szerszy nurt namysłu nad wartością danych, uzyskiwanych w procesie badawczym, a może nawet w refleksję nad tym, w jakim stopniu badania mogą odzwierciedlać rzeczywistość. W głowach metodologów – i to nie tylko tych skłonnych tropić błędy i niedostatki badań socjologicznych – pojawiają się pytania o rzetelność badań, o jakość całego procesu badawczego, o wiarygodność otrzymanych rezultatów. Czy zadając pytania o opinie rzeczywiście docieramy do tego, co nasi respondenci sądzą na dany temat, czy też musimy zadowolić się tylko powierzchownymi, formułowanymi „na poczekaniu” oficjalnymi opiniami, którymi badani zechcą się z nami podzielić? Ten brak pewności co do jakości otrzymanych rezultatów, a przez to i jakości badań socjologicznych skłania do szukania ciągle nowych sposobów oceniania zebranego materiału.
The authors of the book, students of the founders and representatives of the Lodz school of methodology in sociology, reflect on verification analyses, which were the subject of theoretical and empirical researches in that school for many years. Jan Lutyński’s unique conception of analyses for internal verification, which was unique in methodology of survey research, remains a particularly important reference point of those reflections. Attempts to face the legacy of the school along with its discussion and assessment, are accompanied by voices of the school’s representatives referring to the then concepts, research enterprises and achievements from today’s perspective. Interviewing the majority of the school’s researchers enabled to include this element into the retrospective. As a result, methodological issues are intertwined with issues of interest to sociology of science. Continuing research interests of their teachers, the authors explore possibilities of the widely understood internal verification by studying the technique of anchoring vignettes in the second part of the work. This method was developed by King et al. (2004) to address the problem of interpersonal and cross-cultural incomparability, also called differential item functioning (DIF), that may affect respondents’ selfassessments. DIF occurs when respondents use the ordinal response categories in different ways because they differ in their evaluation standards. As a consequence, measuring reality only by asking about respondents’ perception can be misleading. The method can be correct for DIF by supplemental vignette questions that make it possible to construct a common scale of measurement across respondents. However, this approach requires some assumptions for results to be valid. One of them, that is response consistency (RC), says that each respondent uses the response categories for the self-assessment question in the same way that he or she uses them to evaluate hypothetical scenarios in vignettes (one person applies the same type of DIF across these two types of questions). To test this assumption external data (or objective indicators) are needed to compare an object under study with some standard. In our study we investigated whether DIF occurs for material living conditions and needs for lifelong learning. We also aimed to check whether RC assumption holds for these two domains. Data were gathered in 2011 using quota sample (n = 1000) of citizens from the Skierniewice subregion, that is women aged 25–59 and men aged 25–64. The sample was selected by three criteria: gender, age and place of living. According to our findings law-income and high-income respondents do differ in their standards for what constitutes satisficing level of material living conditions, however, DIF was not sizeable. A similar conclusion applies for the second domain under study for respondents practicing and not practicing lifelong learning, respectively. Moreover, for material living conditions we find some evidence that RC assumption can be violated indicating that low-income respondents use different DIF when providing self-assessment and evaluating scenarios in vignettes. The reason we suggest is that low-income respondents found the question about their living conditions threating their good self-esteem. - Dostawca treści:
- Biblioteka Nauki
Książka