Article credentials

Original Research

Religious Language in Suicide-Related Reddit Posts: Emotional Tone (Negativity vs. Positivity) in a LargeScale Secondary Text Analysis

A World Journal of Human Dynamics and Social Relations

ISSN Online: 3028-1997 | Print: 3028-1962

Volume 7 | Issue 1 | 2026 | 160 – 172

Melvin A. Macuha, RPsy, ORCID No. 0000-0003-2748-3586

Psychology Program Chairperson, St. Paul University Manila, 680 Pedro Gil Street, Malate, Manila, Philippines

Article History:

Initial submission: 26 February 2026
First decision: 28 February 2026
Revision received: –
Accepted for publication: 20 March 2026
Online release: 08 April 2026

Quick links

Ready to submit? Click the button below.

Need more help before you submit?

Abstract

Religious and spiritual concerns often occur during suicidal crises; however, little is known whether religious language in suicide-related online discourse tends to be emotionally negative (e.g., despair, guilt, abandonment) or positive (e.g., hope, comfort, meaning). Using a large and publicly available body of Reddit posts drawn from suicide-related communities, this study examined the emotional tone associated with religious language in suicide-related posts. The researcher conducted a secondary analysis of the Suicide and Depression Detection dataset (Kaggle), restricting analyses to posts labeled “suicide” (N= 116,037). Religious language was operationalized using a transparent keyword-based lexicon. Emotional tone was estimated through VADER sentiment scoring. Prevalence estimates were reported and two logistic regression models were tested: (a) whether religious language predicted high negativity and (b) whether religious language predicted high positivity, controlling for post length. Religious language appeared in 10.91% of suicide-labeled posts and was associated with heightened emotional polarity and a significantly more negative overall sentiment. Controlling for post length, religious language predicted greater odds of high negativity (OR = 1.50, 95% CI [1.44, 1.57], p < .001) and lower odds of high positivity (OR = 0.79, 95% CI [0.75, 0.83], p < .001). Findings suggest that, within suicide-related Reddit discourse, religious language is more frequently embedded in negative emotional framing than in positive emotional expression. Ethical safeguards for sensitive public data research are discussed, along with implications for spiritually integrated crisis communication and future cross-cultural studies.

Keywords: Digital mental health, Reddit, Religion, Sentiment analysis, Spirituality, Suicide, VADER

Cite this article

APA (7th edition)

Macuha, M. A. (2026). Religious language in suicide-related Reddit posts: Emotional tone (negativity vs. positivity) in a large-scale secondary text analysis. Social Science Lens: A World Journal of Human Dynamics and Social Relations, 7(1), 160–172. https://doi.org/10.62718/vmca.ssl-wjhdsr.7.1.SC-0226-031.

Author contributions

The researcher conceptualized and designed the study, conducted data management and computational analyses, interpreted the findings, and drafted and revised the manuscript.

Funding

This research received no external funding.

Conflict of interest

The authors no conflict of interest.

Institutional ethics review statement

Ethics approval was not required because the study involved secondary analysis of publicly available, deidentified data with no direct participant interaction.

Data availability statement

The dataset analyzed is publicly available through the Kaggle repository: https://www.kaggle.com/datasets/nikhileswark omati/suicide-watch.

Declaration of generative AI use/assistance

Computational assistance tools were used to support portions of analytic Python coding, specifically ChatGPT (OpenAI, 2026); however, all analytic decisions, model specifications, and interpretations were independently reviewed and verified by the researcher.

Acknowledgement

– (not available).

Publisher’s disclaimer

The views expressed in this article are those of the authors and do not necessarily reflect the views of the publisher. The publisher disclaims any responsibility for errors or omissions.

×

Cart