Relative Serial Recall Accuracy

Measure how well participants maintain local order during retrieval.

The relative serial recall accuracy curve scores a recall as correct when it immediately follows its predecessor in the study list. The first recall is scored relative to position 0, so only study position 1 counts as correct at the first output slot. This measure highlights how well participants maintain local order during retrieval.

Workflow

Code 
import os
import matplotlib.pyplot as plt
import warnings
from jaxcmr.analyses.relative_srac import plot_relative_srac
from jaxcmr.helpers import find_project_root, generate_trial_mask, load_data, save_figure

warnings.filterwarnings("ignore")
Code 
data_path = "data/HealeyKahana2014.h5"
figure_dir = "results/figures"
figure_str = ""
ylim = None
trial_query = "data['listtype'] == -1"
size = 3
confidence_level = 0.95
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_relative_srac(datasets=data, trial_masks=trial_mask, size=size, confidence_level=confidence_level)
if ylim is not None:
    for ax in plt.gcf().axes:
        ax.set_ylim(ylim)
save_figure(figure_dir, figure_str)

Interpretation

The x-axis shows output position and the y-axis shows the probability that the recall at that position immediately follows its study predecessor. Key patterns:

  • High values: participants frequently maintain local study order in their recall sequence.
  • Declining across output positions: local order preservation weakens as the recall sequence progresses.

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.
  • size — maximum number of study positions a single item can occupy.
  • confidence_level — confidence level for subject-wise error bars.