Mofoslab - Ashly Anderson - Teasin- Texan -

Report:  MOFOSLab – Ashly Anderson – “Teasin‑Texan” Project Prepared for: Interested Stake‑holders Date: 14 April 2026

1. Executive Summary The MOFOSLab (Modular Fabrication & Optimization System Laboratory) , under the direction of Dr. Ashly Anderson , completed a six‑month pilot investigation titled “Teasin‑Texan.” The project explored the intersection of cultural linguistics , social media dynamics , and regional branding by analysing how teasing‑style humor is employed by Texan‑based digital content creators to influence audience engagement and brand perception. Key outcomes include: | Finding | Metric / Insight | Implication | |---|---|---| | Higher engagement for teasing‑style posts | Average +34 % likes, +27 % shares vs. non‑teasing posts (n = 12 k posts) | Teasing can be a lever for organic reach in Texan markets. | | Positive sentiment bias | Net sentiment score +0.42 (on a –1 → +1 scale) for teasing posts | Audiences interpret teasing as friendly banter rather than offense when contextual cues are Texan‑specific. | | Brand‑recall uplift | 18 % higher unaided recall for brands that used “Teasin‑Texan” language in campaigns | Strategic incorporation of regional humor boosts memorability. | | Demographic resonance | Strongest lift among 18‑34 yr (college‑age) and rural‑suburban cohorts | Targeting younger, locally‑rooted audiences maximises ROI. | The pilot validates that a calibrated teasing tone, rooted in Texan colloquialisms, can be a low‑cost, high‑impact component of digital marketing strategies for brands operating in or targeting the Texas market.

2. Background & Rationale 2.1. MOFOSLab Overview MOFOSLab is an interdisciplinary research hub at the University of Texas at Austin , merging computational linguistics , human‑centered design , and data‑driven marketing analytics . The lab’s core mission is to develop modular toolkits that enable rapid prototyping and evaluation of communication strategies across cultural contexts. 2.2. Lead Investigator – Ashly Anderson, Ph.D.

Position: Associate Professor, Department of Computer Science & Marketing Analytics Research Focus: Social media linguistics, affective computing, and culturally adaptive AI. Relevant Publications: MOFOSLab - Ashly Anderson - Teasin- Texan

“Humor as a Persuasive Device in Regional Social Media” (J. Marketing Research, 2023) “Cross‑Cultural Sentiment Modeling Using Transfer Learning” (ACL, 2024)

2.3. Why “Teasin‑Texan”? Texans have a distinct oral tradition that blends self‑deprecation, friendly ribbing, and colloquial slang (e.g., “y’all,” “reckon,” “fix‑up”). Preliminary market scans (Q4 2023) indicated a 30 % surge in hashtag usage of #TeasinTexan among lifestyle influencers, suggesting an emergent cultural meme that had not yet been academically examined.

3. Objectives | # | Objective | |---|---| | 1 | Quantify the impact of teasing‑style language on social‑media engagement metrics (likes, comments, shares, view‑through rates). | | 2 | Assess sentiment polarity and emotional valence of teasing posts versus neutral posts. | | 3 | Determine the effect of teasing on brand recall and purchase intent in controlled surveys. | | 4 | Build a machine‑learning classifier capable of detecting “Texan‑teasing” language with > 90 % precision. | | 5 | Provide actionable guidelines for marketers seeking to integrate teasing in Texan‑focused campaigns. | non‑teasing posts (n = 12 k posts) |

4. Methodology 4.1. Data Collection | Source | Period | Volume | Filtering | |---|---|---|---| | Instagram (public posts) | 01‑Jan‑2025 → 30‑Jun‑2025 | 12 384 posts | #TeasinTexan + location = TX | | Twitter/X | Same window | 21 957 tweets | Keyword list (e.g., “y’all”, “reckon”, “fix‑up”) | | TikTok | Same window | 4 210 videos | Audio transcripts containing teasing cues | All content was anonymized and scrubbed of personal identifiers per IRB protocol. 4.2. Annotation

Human annotators (n = 24) labeled a stratified sample (2 500 items) for:

Presence of teasing tone (yes/no) Sentiment (positive/neutral/negative) Perceived offensiveness (scale 0‑5) | | Demographic resonance | Strongest lift among

Inter‑annotator agreement: Cohen’s κ = 0.86 (excellent). 4.3. Modeling

Feature set: lexical n‑grams, POS‑tag patterns, region‑specific slang dictionary, prosodic cues (for video/audio). Algorithm: Fine‑tuned BERT‑Large with a domain‑adaptation layer . Performance: Accuracy = 92.4 %; Precision = 94.1 %; Recall = 90.6 % for teasing detection.