*technically removed people whose number of incorrect moves in final episode (within a series) is higher than in their first episode)

Between vs. within category distances for cel

Simple descriptives and correlations for number of moves and confidence in guesses by guess attributes: would it solve rule and broadness of guess.

People sort Posner & Keelse-style stimuli into two groups using descriptions given by previous participants who were exposed to the same stimuli either with or without novel labels (and then asked to write descriptions for the two categories)

Looked at whether descriptions produced by people exposed to A and B categories wih=th or without labels produced greater accuracy in new people selecting from A and/or B items (+ prototypes)

Analysis of people's shape selection (from 4 categories of novel shape; A,B,C,D) for descriptions of a specific shape category (A,B) written by a prior participant who was exposed to the shapes either with or without novel labels.

Analysis for accuracy of people selecting the correct prototype or non-prototype shapes on basis of a shape category description created by someone exposed with novel labels, or without labels.

Compares how likely participants are to place within-category items with the category prototype, and takes into account the distortion of a given item away from the prototype. Data is sorting solutions for shapes from one in-lab and two online studies.

Looks at whether categoricality in sorting predicts alignment within individuals. Data is sorting solutions for shapes from one in-lab and two online studies.

Compares average alignment across conditions for sorting similarity of shapes in one in-lab task and two online tasks. Includes a baseline alignment covariate to account for chance alignment between sorters due to number of clusters used when sorting.

Calculates and compares within vs. between category distances for one in-lab and two online shape sorting tasks

Match to sample performance across 1 in-lab task and two online tasks

Glmer and plots examining native and non-native speakers' propensity for lexical alignment with native vs. non-native interlocutors

Gets clustering with prototype info for use in "Qualities of prototype clusters" analysis in CEL_cluster_composition script.

Uses pamk from fpc package to obtain clusters for participants' sorting solutions across conditions.

Get all item pair distances and compare within versus between category distances (for pre-defined a/b category structure) across different label/exposure conditions.

Get compositions of sorters' clusters from pamk data. Looks at cluster purity, size and effects of item distortion - focusing on clusters containing the A or B prototype

This script gets the correct participants' (from the free sort) data for the match-to-sample task (no vs with labels conditions only). We analyze accuracy and RT by condition first, and then focus on RT as a function of sample vs foil difference in disortion (from prototypes).

Getting pairwise alignment for every pair within the conditions of CEL. Can either analyze all binary data with glmer, or bootstrapping approach: average by participant pairs and subset to 100 random pairs each run.

Clustering and by-item pairs probability (of being in same cluster) for sort and label task.

Intraclass correlations for different conditions of CEL study (categorization and labelling effects on category alignment).