Practical Examples¶

This guide provides hands-on examples you can try in PhonoLex to understand each tool's capabilities.

Custom Word Lists Examples¶

Example 1: Simple CVC Words for Early Intervention¶

Goal: Find simple consonant-vowel-consonant words for early speech therapy

Steps: 1. Open Custom Word Lists 2. Add filter: Syllables = 1 3. Add filter: Phonemes ≤ 4 4. Add filter: WCM ≤ 3 (low complexity) 5. Add filter: Frequency ≥ 10 (common words) 6. Generate

Expected Results: - Words like: cat, dog, bed, cup, hat, sit, run - All monosyllabic, simple structure, commonly used

Why This Works: - Syllable count = 1 ensures no multisyllabic words - Phoneme count ≤ 4 keeps words short - Low WCM means fewer complex features (clusters, velars, etc.) - High frequency ensures familiar, functional words

Example 2: Initial /s/ Words with Semantic Scaffolding¶

Goal: Find /s/ initial words that are highly imageable for articulation practice

Steps: 1. Open Custom Word Lists 2. Add pattern: STARTS_WITH → /s/ 3. Add filter: Imageability ≥ 5.0 (highly imageable) 4. Add filter: Frequency ≥ 5 (reasonably common) 5. Generate

Expected Results: - Words like: sun, snake, sock, sand, soap, snow - All start with /s/, easy to visualize

Why This Works: - STARTS_WITH /s/ targets specific phoneme position - High imageability helps children create mental images - Frequency filter ensures words are useful in daily life

Example 3: Late-Developing Sounds in Simple Contexts¶

Goal: Find words with /ɹ/ (r-sound) in simple, early-acquired contexts

Steps: 1. Open Custom Word Lists 2. Add pattern: CONTAINS → /ɹ/ 3. Add filter: Syllables = 1 4. Add filter: MSH ≤ 4.0 (lower motor complexity) 5. Add filter: Age of Acquisition ≤ 4.0 (learned early) 6. Generate

Expected Results: - Words like: hear, hair, air, fair, bear, chair, fear - Late phoneme (/ɹ/) in otherwise simple words

Why This Works: - CONTAINS /ɹ/ finds all positions (initial, medial, final) - Monosyllabic keeps structure simple - Low MSH means other phonemes are early-developing - Low AoA ensures words are familiar to children

Example 4: Negative Valence Words for Emotional Language¶

Goal: Find words with negative emotional content for social-emotional work

Steps: 1. Open Custom Word Lists 2. Add filter: Valence ≤ 3.0 (negative valence) 3. Add filter: Arousal ≥ 5.0 (emotionally activating) 4. Add filter: Frequency ≥ 10 (common words) 5. Add filter: Imageability ≥ 4.0 (can be visualized) 6. Generate

Expected Results: - Words like: afraid, scared, mad, evil, dangerous, attack - Negative emotional tone, high arousal, common usage

Why This Works: - Low valence targets negative emotions - High arousal ensures emotionally salient words - Frequency + imageability make words functional and learnable

Example 5: Excluding Problematic Phonemes¶

Goal: Find /k/ words without /g/ (for children substituting k→g)

Steps: 1. Open Custom Word Lists 2. Add pattern: STARTS_WITH → /k/ 3. Add exclusion: Exclude phoneme → /g/ 4. Add filter: Syllables ≤ 2 5. Generate

Expected Results: - Words like: cat, car, cut, candy, carrot - All have initial /k/, none contain /g/

Why This Works: - STARTS_WITH /k/ targets the sound we're working on - Exclusion prevents words with the substitution pattern - Syllable limit keeps words manageable

Contrastive Sets Examples¶

Example 6: Minimal Pairs - Fronting (k→t)¶

Goal: Generate minimal pairs for a child who fronts /k/ to /t/

Steps: 1. Open Contrastive Sets → Minimal Pairs 2. Target phoneme: /k/ 3. Substitute phoneme: /t/ 4. Position: Initial 5. Word length: Short 6. Complexity: Low 7. Generate

Expected Results: Minimal pairs like: - cat / tat (if "tat" exists) - key / tea - came / tame - cold / told

Why This Works: - Contrasts exactly one phoneme (/k/ vs /t/) - Initial position targets the error position - Short, simple words make practice easier

Example 7: Maximal Opposition - Major Class Difference¶

Goal: Find maximal opposition pairs for a child with moderate-severe SSD

Unknown phonemes: /s, ʃ, θ, l, ɹ/

Steps: 1. Open Contrastive Sets → Maximal Opposition 2. Enter unknown phonemes: /s, ʃ, θ, l, ɹ/ (comma-separated) 3. Click Generate Phoneme Pairs

Expected Results:

Top-ranked pairs (system calculates automatically):

Pair	Major Class?	Feature Differences	Score
/θ/ - /l/	✓ Yes (obstruent-sonorant)	15	115
/s/ - /l/	✓ Yes (obstruent-sonorant)	14	114
/ʃ/ - /ɹ/	✓ Yes (obstruent-sonorant)	13	113

Why This Works: - Major class difference (obstruent vs. sonorant) earns +100 bonus - Many feature differences maximize phonological contrast - Research shows this promotes system-wide learning

Next Step: Select /θ/ - /l/ pair, then click Generate Word Lists to get minimal pairs like: - thin / Lynn - think / link - mouth / mole (word-final)

Example 8: Multiple Opposition - Stopping¶

Goal: Treat a child who produces all of {/s, ʃ, θ, f/} as [t]

Steps: 1. Open Contrastive Sets → Multiple Opposition 2. Substitute phoneme: /t/ (what child says) 3. Target phonemes: /s, ʃ, θ, f/ (what should be said) 4. Click Generate Sets

Expected Results:

Minimal sets (4-word sets): - tie / sigh / shy / thigh / fie - tore / sore / shore / Thor / four - tee / see / she / thee / fee

Why This Works: - Addresses entire phonological collapse pattern simultaneously - Maximal Classification algorithm selects representative targets - Maximal Distinction algorithm finds minimal sets that contrast all targets

Phonological Similarity Examples¶

Example 9: Finding Perfect Rhymes¶

Goal: Find words that rhyme with "cat" for phonological awareness activities

Steps: 1. Open Phonological Similarity 2. Enter target word: cat 3. Select preset: Rhymes (onset=0.0, nucleus=0.5, coda=0.5) 4. Threshold: 0.85 (high similarity) 5. Limit: 20 results 6. Click Find Similar Words

Expected Results: - bat (0.95), hat (0.94), sat (0.93), mat (0.92), rat (0.91), pat (0.90) - All share /æt/ ending (nucleus + coda)

Why This Works: - Zero onset weight ignores initial consonants - High nucleus/coda weights prioritize vowel and final consonant match - High threshold (0.85) returns only near-perfect rhymes

Example 10: Finding Alliteration¶

Goal: Find words that alliterate with "snap" for phonological awareness

Steps: 1. Open Phonological Similarity 2. Enter target word: snap 3. Select preset: Alliteration (onset=1.0, nucleus=0.0, coda=0.0) 4. Threshold: 0.70 (moderate similarity) 5. Limit: 20 results 6. Click Find Similar Words

Expected Results: - Words starting with /sn/, /s/, or similar clusters: - snake (0.85), snack (0.83), snow (0.80) - slip, slide, slow (lower scores, similar onset)

Why This Works: - Onset weight = 1.0 prioritizes initial sounds - Zero nucleus/coda weights ignore vowels and endings - Moderate threshold allows some variation in cluster structure

Example 11: Custom Weights - Consonant Frames¶

Goal: Find words with similar consonant frame to "cat" (ignore vowel)

Steps: 1. Open Phonological Similarity 2. Enter target word: cat 3. Select preset: Consonance (onset=0.5, nucleus=0.0, coda=0.5) 4. Adjust threshold: 0.75 5. Limit: 15 results 6. Click Find Similar Words

Expected Results: - kit (0.82), cut (0.78), cot (0.78), coat (0.76) - All have /k_t/ or similar consonant frame

Why This Works: - Zero nucleus weight completely ignores vowel differences - High onset/coda weights prioritize consonant matches - Useful for identifying phoneme confusions based on consonant context

Lookup Examples¶

Example 12: Word Properties Lookup¶

Goal: Get complete phonological and psycholinguistic profile for "strength"

Steps: 1. Open Lookup → Word Lookup 2. Enter: strength 3. View results

Expected Results:

Word: strength
IPA: /strɛŋkθ/
Syllables: 1
Phonemes: 7

Phonological Complexity:
- WCM: 11 (very high - 3-consonant cluster, velars, fricatives)
- MSH: 5.5 (high motor complexity - fricatives and velar nasal)

Lexical:
- Frequency: 28.5 (fairly common)
- AoA: 5.8 (learned later)

Semantic:
- Imageability: 3.2 (abstract concept)
- Familiarity: 6.1 (familiar word)
- Concreteness: 2.5 (abstract)

Affective:
- Valence: 6.8 (positive)
- Arousal: 5.2 (moderately arousing)
- Dominance: 7.1 (high dominance/power)

Why This is Useful: - High WCM/MSH indicates late-developing word structure - Abstract (low imageability/concreteness) makes it harder for children - Late AoA aligns with complexity - Positive valence and high dominance fit semantic associations with "power"

Example 13: Phoneme Feature Comparison¶

Goal: Understand why /t/ and /d/ are a minimal pair (voice contrast)

Steps: 1. Open Lookup → Phoneme Comparison 2. Phoneme 1: /t/ 3. Phoneme 2: /d/ 4. Click Compare

Expected Results:

Shared Features (36 matching): - consonantal: + - sonorant: - - continuant: - - nasal: - - labial: - - coronal: + - anterior: + - dorsal: - - (etc.)

Different Features (2): - voice: /t/ = - (voiceless) | /d/ = + (voiced) - periodicGlottalSource: /t/ = - | /d/ = +

Similarity Score: 0.947 (very similar - only voice differs)

Why This is Useful: - Shows exactly which features distinguish minimal pairs - Explains why /t/ - /d/ is a common substitution (only 1 major feature differs) - Demonstrates that voiced/voiceless pairs share almost all features

Example 14: Finding Phonemes by Features¶

Goal: Find all fricatives for articulation hierarchy planning

Steps: 1. Open Lookup → Search by Features 2. Add feature: consonantal = + 3. Add feature: sonorant = - 4. Add feature: continuant = + 5. Click Search

Expected Results:

Phonemes found (9): - Voiceless fricatives: /f, θ, s, ʃ, h/ - Voiced fricatives: /v, ð, z, ʒ/

Why This is Useful: - Identifies entire sound classes systematically - Helps plan treatment hierarchies (fricatives often later-developing) - Shows relationships between sounds (voice pairs)

Example 15: Comparing Maximal Opposition Phonemes¶

Goal: Verify that /s/ and /l/ differ maximally (obstruent vs. sonorant)

Steps: 1. Open Lookup → Phoneme Comparison 2. Phoneme 1: /s/ 3. Phoneme 2: /l/ 4. Click Compare

Expected Results:

Similarity Score: 0.632 (very different - 14 features differ)

Major Class Difference: YES (sonorant: - vs +)

Why This is Useful: - Confirms maximal opposition (major class + many features) - Explains why /s/ - /l/ is effective for intervention - Shows that even though both are coronal and anterior, they differ greatly

Advanced Queries¶

Example 16: High-Frequency Abstract Words¶

Goal: Find abstract words for older children working on vocabulary

Steps: 1. Open Custom Word Lists 2. Add filter: Frequency ≥ 20 (high frequency) 3. Add filter: Concreteness ≤ 2.5 (abstract) 4. Add filter: Age of Acquisition ≥ 5.0 (learned later) 5. Generate

Expected Results: - Words like: justice, theory, particular, professional, affair, destiny, former, instance - Common abstract concepts learned in later school years

Example 17: Emotionally Neutral Words¶

Goal: Find emotionally neutral words for baseline testing

Steps: 1. Open Custom Word Lists 2. Add filter: Valence = 4.0 - 6.0 (neutral range) 3. Add filter: Arousal = 3.0 - 5.0 (low arousal) 4. Add filter: Frequency ≥ 10 5. Generate

Expected Results: - Words like: table, paper, floor, time, thing, work, wait, put - Everyday words without emotional content

Example 18: Multi-Pattern Complex Query¶

Goal: Find words with initial /k/ AND final /t/, excluding /s/

Steps: 1. Open Custom Word Lists 2. Add pattern: STARTS_WITH → /k/ 3. Add pattern: ENDS_WITH → /t/ 4. Add exclusion: Exclude phoneme → /s/ 5. Add filter: Frequency ≥ 5 6. Generate

Expected Results: - Words like: cat, kit, court, caught, coat - All start with /k/, end with /t/, contain no /s/

Why This is Useful: - Demonstrates AND logic (must match ALL patterns) - Exclusions refine word lists for specific therapy goals - Complex queries help find exactly the right stimuli

Tips for Using Examples¶

Start Simple: Try examples 1-4 first to understand basic filtering
Experiment: Adjust thresholds and limits to see how results change
Compare Tools: Try the same goal across different tools (e.g., finding rhymes in Custom Word Lists vs Phonological Similarity)
Export Results: Use CSV export to save word lists for therapy materials
Combine Filters: Use multiple filters to create highly specific lists

Next Steps¶

Review Custom Word Lists for detailed feature explanations
Read Technical Architecture to understand how similarity scores are calculated
Check Psycholinguistic Norms Reference for complete property descriptions