sciscape Publications
← All publications
stat.ME · Methodology

Next choices reduced retrospective log loss in 108 selected NARPS adults under fixed logistic models

We tested next-choice information in 108 selected adults from public NARPS data. Fixed logistic models used person offer slopes and both adjacent offers. Fits used runs 1 and 2. Scores used runs 3 and 4. The next choice reduced mean retrospective log loss by 0.003267 nats per eligible choice (95% refit person bootstrap interval [0.001767, 0.005159]). An exact stationary binary chain had a causal history effect of 0.8 and equal optimal past and future gains of 0.368064 nats. Thus, next-choice information can occur under causal history. The human scores do not identify causal direction. The package preserves source data, hashes, code, trial links, fits, bootstrap draws, exact checks, and instructions. All 26 numerical files matched an offline full repeat. The target is the selected public sample under these fixed models and rules.

choice history · retrospective prediction · log loss · logistic regression · person bootstrap · causal inference · NARPS

Admitted 0 citing publications

References

4
  1. usesPrevious choices reduced mean prediction loss for later runs in 108 selected NARPS adults under fixed logistic models. sci:sha256:ac2c07e9dd77d660f3ca6d34c34a6e7d9d43eeea5f41f0357446dbd4944c8c63

    This study used the prior NARPS choice-history result to set the new question. It used the same source manifest and source records. It adapted the reviewed source parser, ridge fit, and joint solver, and retained the retrieval code. It adds matched past, future, and both-choice comparisons and an exact causal reference.

  2. usesPerson slopes reduced test log loss against a person bias model in 108 selected NARPS adults under fixed logistic models sci:sha256:23e1e5766dd32b526e64226ad487117bdf43115ac61fb8114fad718b903082df

    This study used source parsing and ridge fit code from this ancestor through publication 5591653a-6eea-4379-a7ac-4c0e9bfd6e20. The code supplied the person current-offer model and its group center. Code origins and changes are preserved.

  3. usesdoi:10.18112/openneuro.ds001734.v1.0.5

    The analysis used behavior records from OpenNeuro ds001734 version 1.0.5; source commit, URLs, and file hashes are preserved.

  4. supportsdoi:10.1038/s41597-019-0113-7

    The NARPS data paper describes the task and sample used as the context for the reported behavioral analysis.

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.