Coded quesitonnaire data on urban noise sources, conceptions, and coping in Indonesian Cities

Published: 20 October 2026| Version 2 | DOI: 10.17632/32g53jp2d6.2
Contributors:
,

Description

This dataset contains coded data derived from open-ended questionnaire responses by 145 residents of Indonesian cities about the noise that disturbs them most, how they conceive of it, and how they respond to it. Data were collected online in Bahasa Indonesia between 15 March and 26 April 2026. Responses were segmented into meaning units and coded into three blocks: Noise Source (3 categories), Noise Conception (6 categories), and Noise Solution (5 categories). Each category is provided as a respondent-level binary indicator (1 = at least one mention). Items on the location, activity, and time of noise served as recall anchors and were not analysed. The dataset supports the article "The individualisation of noise: decoupled through appraisal, coupled through environment in urban Indonesia". The coded data were analysed through multiple correspondence analysis (MCA): three pairwise analyses (Source × Conception, Conception × Solution, Source × Solution), two sensitivity analyses of Source × Solution, and a joint analysis of all 14 categories. Files: coded_matrix.csv (respondent-level indicator matrix with age band, activity, location type and time of noise), codebook.xlsx (categories, sub-categories, codes, and definitions), instrument.docx (questionnaire in Indonesian with English translation), mca_reproduction.R (script reproducing all analyses), and README.txt (file structure and variable dictionary). Raw verbatim responses are not included because participants were told their answers would be kept confidential. No personally identifying information is included.

Files

Steps to reproduce

1. Instrument. An open-ended questionnaire in Bahasa Indonesia asked about the noise source, how respondents felt and thought about it, and how they avoided or coped with it. Items on place, activity and time of noise were asked first as recall anchors (Tulving and Thomson 1973; Tourangeau et al. 2000) and were not analysed. See instrument.docx. 2. Data collection. Online questionnaire (Google Forms), 15 March to 26 April 2026, non-probability snowball sampling; 145 complete responses from residents of Indonesian cities. 3. Coding. Responses were segmented into meaning units and coded by two researchers through staged consensus into three blocks: Noise Source (3 categories), Noise Conception (6) and Noise Solution (5). Labels were theory-derived (Lazarus and Folkman 1984; Basner et al. 2014) or inductive (Saldaña 2021); see codebook.xlsx. 4. Indicator matrix. Each category was converted into a respondent-level binary indicator (1 = at least one segment in that category), giving coded_matrix.csv. 5. Analysis. Multiple correspondence analysis on the indicator matrix, each binary variable entering with two levels: MCA 1 (Source x Conception, age group supplementary), MCA 2 (Conception x Solution), MCA 3 (Source x Solution), MCA 3 without Mobility Strategy, MCA 3 without Mobility Strategy and Policy & Urban Scale Solution, and a joint MCA of all 14 categories. Explained variance is reported as Greenacre-adjusted inertia. A category was interpreted on a dimension when the contributions of its two levels, summed, were at least 1/Q (Q = number of active variables) and its squared cosine was at least 0.20. 6. Software and reproduction. Analyses were run in JMP Pro Student 19.0, which reports Greenacre-adjusted inertia, contributions and squared cosines directly. Running "Rscript mca_reproduction.R coded_matrix.csv" (base R, no packages) reproduces all six analyses with values identical to the published tables after rounding; results are written to mca_results/.

Institutions

Categories

Public Health, Noise in the Environment, Environment and Health, Qualitative Research, Coping, Multiple Correspondence Analysis

Licence