Izzivi spletnih verjetnostnih panelov: primer panela CRONOS : primer panela CRONOS

Šoštarič, Tilen (author), Vehovar, Vasja (mentor)

V zadnjih letih zaradi naraščajoče rabe interneta spletno anketiranje izpodriva ostale načine anketiranja. V tem okviru predstavlja enega največjih izzivov vprašanje, ali lahko verjetnostni spletni paneli nadomestijo osebno terensko anketiranje, saj se lahko anketiranci, ki se odločijo za sodelovanje, razlikujejo od oseb, ki v panelu ne sodelujejo. V diplomskem delu smo preučili pristranskost ocen, ki pri tem nastanejo zaradi neodgovora in nepokritja. Empirično raziskavo smo vezali na Evropsko družboslovno raziskavo (ESS), ki je leta 2017 (ESS, Round 8) izvajala eksperiment CRONOS, v katerem so v treh državah (Estonija, Slovenija in Velika Britanija) respondente v ESS prosili za nadaljnje sodelovanje v spletnem panelu. Preverili smo, ali se odgovori respondentov, ki so se odločili za sodelovanje v panelu, razlikujejo od odgovorov respondentov, ki se za sodelovanje niso odločili. Osredotočili smo se zgolj na podatke ESS, kjer smo primerjali vzorec ustreznih respondentov ESS s podvzorcem ESS respondentov, ki so kasneje sodelovali v prvem valu panela CRONOS, saj bi lahko pri primerjanju ESS in podatkov iz panela CRONOS prišlo do razlik zaradi drugačnega načina anketiranja (ESS osebno terensko anketiranje, CRONOS spletno anketiranje). Ugotovili smo, da je pristranskost zaradi nepokritja in neodgovora kljub naraščajoči uporabi interneta pri številnih spremenljivkah še vedno prisotna.

• Language: Slovenian
• Keywords: CRONOS, Evropska družboslovna raziskava, spletni paneli, verjetnostno vzorčenje
• Work type: Bachelor thesis/paper
• Typology: 2.11 - Undergraduate Thesis
• Organization: FDV - Faculty of Social Sciences
• Place of publishing: Ljubljana
• Publisher: [T. Šoštarič]
• Year: 2020
• Number of pages: 74 str.
• PID: 20.500.12556/RUL-120475
• UDC: 303(043.2)
• COBISS.SI-ID: 40478979
• Publication date in RUL: 20.09.2020

Secondary language

• Language: English
• Title: Challenges of probability-based online panels: CRONOS panel : diplomsko delo

Abstract

In recent years, due to the growing use of the Internet, web surveys have been increasingly replacing other survey modes. One of the biggest challenges in this context is the question of whether probability-based online panels are able to replace face-to-face field surveys, as respondents who choose to participate in online surveys may differ from those who do not participate in such panels. This diploma thesis focuses on the bias of the estimates that occurs due to nonresponse and noncoverage in a probability-based online panels. The empirical study was based on European Social Survey (ESS, Round 8) where CRONOS experiment was conducted in 2017. The experiment asked ESS respondents from three countries (Estonia, Slovenia and the United Kingdom) to participate in the follow-up online panel. The ESS data were analysed to determine whether the participants in the panel differed from non-participants. The research focused solely on the ESS where we compared the entire sample of eligible respondents to subsample of respondents who participated in the first wave of CRONOS panel, as the comparison of data between the ESS and CRONOS panel could lead to additional differences due to different survey mode (ESS face-to-face, CRONOS web). We discovered that bias due to noncoverage and nonresponse was still present in many variables despite the increasing use of the Internet.

• Keywords: CRONOS, European Social Survey, online panels, probability sampling