context-framed-listening / docs

Framed Listening — Analysis Documentation

van der Walle, H. A. · Durham University Music Psychology · 2026

Setup

Setup Descriptive stats before splitting trials

Descriptive Statistics

Descriptive Statistics MIMC description word counts

Text Preprocessing

Text Preprocessing Spell checking NLP Stop words NLP Lemmatise NLP Apply text preprocessing Creating combined combMIMC output

Setup

Setup Python Load Data Helper Functions

Group A — Aggregated combMIMC (N = 64)

A1 — TF-IDF TF-IDF A2 — Word2Vec (lvl2a) W2V A3 — Word2Vec (lvl1a) W2V A4 — BERT BERT A5 — BERTweet BERT A6 — paraphrase-MiniLM-L6-v2 SBERT A7 — all-MiniLM-L6-v2 SBERT A8 — all-mpnet-base-v2 SBERT A9 — paraphrase-mpnet-base-v2 SBERT Group A — Cross-Model Comparison summary

Group B — Individual MIMC (N = all)

B1 — TF-IDF TF-IDF B2 — Word2Vec (lvl2a) W2V B3 — Word2Vec (lvl1a) W2V B4 — BERT BERT B5 — BERTweet BERT B6 — paraphrase-MiniLM-L6-v2 SBERT B7 — all-MiniLM-L6-v2 SBERT B8 — all-mpnet-base-v2 SBERT B9 — paraphrase-mpnet-base-v2 SBERT Group B — Cross-Model Comparison summary Summary

Setup & Data

Setup Python Data Preparation Load Data Sentence Embeddings BERT Pairwise Similarity by Condition cosine

Linear Mixed Models

Linear Mixed Models LMM LMM setup & prep LMM Models Supplementary Study — believability moderator supp. Estimated Marginal Means

Research Questions

RQ1 — Context shape MIMC similarity? RQ1 Omnibus Test Context effect: dc_SX_dg vs dc_dx_dg Visualisation plot RQ2 — Genre moderate context effect? RQ2 Framing susceptibility by genre Within-genre clip vs context dominance Visualisation plot RQ3 — Clip or context drives convergence? RQ3 Clip vs Context (RQ3a) RQ3b — Combined advantage SC_SX Visualisation plot

Wrap-up

Summary Exploratory Analyses Pairwise context & genre comparisons Similarity matrices Individual-level t-SNE

Setup & Data

Setup Python Data Preparation Load Data TF-IDF Vectorisation TF-IDF Cosine Similarity Matrix cosine Extract similarity values by condition

Linear Mixed Models

Linear Mixed Models LMM LMM setup & prep LMM Models Estimated Marginal Means

Research Questions

RQ1 — Context shape MIMC similarity? RQ1 Omnibus Test Context effect: dc_SX_dg & dc_SX_SG Visualisation plot RQ2 — Genre moderate context effect? RQ2 Framing susceptibility by genre Within-genre clip vs context dominance Visualisation plot Supplementary Study — believability moderator supp. RQ3 — Clip or context drives convergence? RQ3 Clip vs Context (RQ3a) RQ3b — Combined advantage SC_SX Visualisation plot

Wrap-up

Summary Exploratory Analyses Word clouds by genre × context Pairwise context & genre comparisons Similarity matrices Document-level t-SNE

No sections match.