Continuous Estimation of a Time-Varying Probability

Robinson, Gordon H.

doi:10.1080/00140136408930721

Cited by 38 publications

(61 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Notable exceptions might be found in prior work on regime shifts (Massey & Wu, 2005), change detection in nonstationary systems (Brown & Steyvers, 2009;Chinnis & Peterson, 1970) and in an autocorrelation paradigm (Speekenbrink, Twyman, & Harvey, 2012), or continuous probability updating (Rapoport, Stein, & Burkheimer, 1979;Robinson, 1964). However, no prior research has investigated inferences about latent changes in proportions under orthogonally manipulated changes in the size of stimulus samples.…”

Section: Detection Of Changementioning

confidence: 90%

Anomalies in the detection of change: When changes in sample size are mistaken for changes in proportions

et al. 2015

View full text Add to dashboard Cite

Detecting changes, in performance, sales, markets, risks, social relations, or public opinions, constitutes an important adaptive function. In a sequential paradigm devised to investigate detection of change, every trial provides a sample of binary outcomes (e.g., correct vs. incorrect student responses). Participants have to decide whether the proportion of a focal feature (e.g., correct responses) in the population from which the sample is drawn has decreased, remained constant, or increased. Strong and persistent anomalies in change detection arise when changes in proportional quantities vary orthogonally to changes in absolute sample size. Proportional increases are readily detected and nonchanges are erroneously perceived as increases when absolute sample size increases. Conversely, decreasing sample size facilitates the correct detection of proportional decreases and the erroneous perception of nonchanges as decreases. These anomalies are however confined to experienced samples of elementary raw events from which proportions have to be inferred inductively. They disappear when sample proportions are described as percentages in a normalized probability format. To explain these challenging findings, it is essential to understand the inductivelearning constraints imposed on decisions from experience.

show abstract

Section: Detection Of Changementioning

confidence: 90%

Anomalies in the detection of change: When changes in sample size are mistaken for changes in proportions

et al. 2015

View full text Add to dashboard Cite

show abstract

“…The weight of evidence is highest when diagnosticity is high and the system is unstable (i.e., high transition probability), and lowest when diagnosticity is low and the system is stable (i.e., low (1984), and Robinson (1964). transition probability).…”

Section: Research Hypothesismentioning

confidence: 99%

“…The studies have varied considerably in methodology. In some studies, the sequences were presented sequentially (Robinson, 1964;Chinnis & Peterson, 1968, 1970Barry & Pitz, 1979), while in other studies, a sequence was provided in its entirety (Theios et al, 1971). Some subjects estimated when the process changed from one regime to another (Barry & Pitz, 1979;Theios et al, 1971), whereas other subjects estimated the probability that the data were drawn from one regime or another (Chinnis & Peterson, 1968, 1970.…”

Section: Background Previous Researchmentioning

confidence: 99%

“…Some subjects estimated when the process changed from one regime to another (Barry & Pitz, 1979;Theios et al, 1971), whereas other subjects estimated the probability that the data were drawn from one regime or another (Chinnis & Peterson, 1968, 1970. Studies have also varied environmental parameters such as diagnosticity (Robinson, 1964;Chinnis & Peterson, 1968, 1970, the payoff structure (Barry & Pitz, 1979), and the rate at which subjects receive stimuli (Robinson, 1964).…”

Section: Background Previous Researchmentioning

confidence: 99%

See 1 more Smart Citation

Detecting Regime Shifts: The Causes of Under- and Over-Reaction

Massey

2004

SSRN Journal

View full text Add to dashboard Cite

Many decision makers operate in dynamic environments in which markets, competitors, and technology change regularly. The ability to detect and respond to these regime shifts is critical for economic success. We conduct three experiments to test how effective individuals are at detecting such regime shifts. Specifically, we investigate when individuals are most likely to underreact to change and when they are most likely to overreact to it. We develop a system-neglect hypothesis: Individuals react primarily to the signals they observe and secondarily to the environmental system that produced the signal. The experiments, two involving probability estimation and one involving prediction, reveal a behavioral pattern consistent with our systemneglect hypothesis: Underreaction is most common in unstable environments with precise signals, and overreaction is most common in stable environments with noisy signals. We test this pattern formally in a statistical comparison of the Bayesian model with a parametric specification of the system-neglect model. ABSTRACTMany decision makers operate in dynamic environments, in which markets, competitors, and technology change regularly. The ability to detect and respond to these regime shifts is critical for economic success. We conduct three experiments to test how effective individuals are at detecting such regime shifts. Specifically, we investigate when individuals are most likely to under-react to change and when they are most likely to over-react to it. We develop a systemneglect hypothesis: individuals react primarily to the signals they observe and secondarily to the environmental system that produced the signal. Three experiments, two involving probability estimation and one involving prediction, reveal a behavioral pattern consistent with our systemneglect hypothesis: under-reaction is most common in unstable environments with precise signals and over-reaction is most common in stable environments with noisy signals. We test this pattern formally in a statistical comparison of the Bayesian model with a parametric specification of the system-neglect model.

show abstract

“…Conversely, when an observer detects too few change points, they may fail to adapt to shortlived changes in the environment. This tradeoff between detecting too few and too many change points has often been ignored in previous studies of change detection, which mostly assumed an offline experiment where the task is to identify change points in a complete set of data that were observed earlier (see, e.g., Chinnis & Peterson, 1968, 1970Massey & Wu, 2005;Robinson, 1964). However, real-world examples are invariably online: data arrive sequentially, and a detection response is required as soon as possible after a change point passes, before all the data have been observed.…”

Section: Introductionmentioning

confidence: 99%