Analisis Komparatif Konsistensi Respons pada Mode Survei Online dan Offline

Abstract

Survei online dianggap lebih efisien karena dapat menjangkau wilayah yang luas, kecepatan penyebaran kuesioner, kemudahan akses data, dan ekonomis. Namun keunggulan ini belum menjamin data yang diperoleh berkualitas (valid, reliabel serta representatif). Penelitian ini berupaya mengungkap tingkat konsistensi data hasil survei online dan offline pada kelompok responden yang sama agar hasilnya dapat dibandingkan dan menghindarkan bias sampling dan bias non respon. Hasil penelitian menunjukkan, tingkat konsistensi jawaban secara keseluruhan dari pertanyaan utama berskala Likert 5 pilihan mencapai 58,9%, dan meningkat menjadi 70,7% setelah dikonversikan menjadi 3 skala pilihan. Tingkat konsisten jawaban biner mencapai 89,6%. Jawaban pertanyaan afeksi relatif lebih konsisten dibanding kognisi. Rata-rata jumlah kata jawaban pertanyaan terbuka survei offline 1,5 kali lebih banyak dari online. Penyebab inkonsistensi antara lain suasana psikologis serta rasa tanggungjawab, kondisi lingkungan responden saat menjawab pertanyaan.

Online surveys are considered to be more efficient because they can reach wider areas, enable faster questionnaire distribution, facilitate easier data access, and require relatively lower costs. However, these advantages do not necessarily guarantee that the data obtained are high quality (validity, reliability, and representativeness). This study seeks to examine the level of consistency between data collected through online and offline surveys from the same group of respondents, allowing the results to be compared while minimizing sampling bias and non-response bias. The findings show that the overall consistency rate for the main questions using five-point Likert scale reached 58.9%, increasing to 70.7% after the responses were recoded into a three-point scale. The consistency rate for binary questions reached 89.6%. Affective questions were found to be relatively more consistent than cognitive ones. The average word count of responses to open ended questions in the offline survey was 1.5 times higher than in the online. The inconsistencies may be attributed to respondents’ psychological state, sense of responsibility, and environmental conditions while answering the questionnaire.