Compound Cueing CRP

Test whether composite or instance-based retrieval governs repeated-item transitions.

The compound cueing analysis tests a prediction that differentiates composite memory models (CMR) from instance-based models (ICMR). For a repeated item at study positions \(i\) and \(j\) (with sufficient spacing), the analysis examines how the two most recent recalls influence the probability of transitioning to the repeated item.

Pure cueing: last two recalls are \(\{i\!-\!2,\, i\!-\!1\}\) or \(\{j\!-\!2,\, j\!-\!1\}\) (both from one occurrence’s neighborhood)
Mixed cueing: last two recalls are \(\{j\!-\!2,\, i\!-\!1\}\) or \(\{i\!-\!2,\, j\!-\!1\}\) (one from each occurrence’s neighborhood)

CMR (\(\tau = 1\)) predicts mixed cueing provides equal or greater support because similarities sum linearly. ICMR (\(\tau > 1\)) predicts pure cueing provides greater support because sharpening amplifies single-trace matches.

Workflow

Code

import os
import matplotlib.pyplot as plt
import warnings
from jaxcmr.analyses.compound_cueing_crp import plot_compound_cueing
from jaxcmr.helpers import find_project_root, generate_trial_mask, load_data, save_figure

warnings.filterwarnings("ignore")

Code

data_path = "data/LohnasKahana2014.h5"
figure_dir = "results/figures"
figure_str = ""
ylim = None
trial_query = 'data["list_type"] > 2'
min_spacing = 6
size = 2

Code

project_root = find_project_root()
figure_dir = os.path.join(project_root, figure_dir)
data_path = os.path.join(project_root, data_path)
data = load_data(data_path)
trial_mask = generate_trial_mask(data, trial_query)

Code

plot_compound_cueing(data, trial_mask, min_spacing=min_spacing, size=size)
if ylim is not None:
    for ax in plt.gcf().axes:
        ax.set_ylim(ylim)
save_figure(figure_dir, figure_str)

Interpretation

The plot compares transition probabilities under pure and mixed cueing conditions. Key patterns:

Mixed >= Pure: consistent with composite retrieval (CMR), where cues from both occurrences sum to support recall.
Pure > Mixed: consistent with instance-based retrieval (ICMR), where concentrated cues from a single occurrence are amplified by sharpening.

API Details

Notebook parameters

data_path — path to an HDF5 file containing a RecallDataset.
figure_dir — directory for saving figures.
figure_str — base filename for the saved figure. Leave empty to display without saving.
ylim — y-axis limits as a tuple, or None for automatic scaling.
trial_query — a Python expression evaluated against the dataset to select trials.
min_spacing — minimum number of intervening items between repeated presentations.
size — maximum number of study positions a single item can occupy.