Contrast effects on sequence assessments - a thesis

Abstract

Loewenstein and his colleagues found divergent preferences for outcomes assessed in isolation versus those embedded in a sequence, i.e. discounting isolated future outcomes versus preferences for increasing and constant sequences. They also found long intervals (i.e. the difference between time delays) rather than long time delays (i.e. the temporal distance from the present) had a detrimental effect on preference for improvement. This thesis proposes a descriptive model of sequence preferences, namely the contrasts model, which acknowledges the difference between interval and delay. The idea is that delay and interval are two different kinds of variables. Delay is non-relational and describes characteristics of individual outcomes, whereas interval is relational and describes characteristics of outcomes in relation to one another. Built on this idea, the contrasts model assumes that the value of a sequence consists of a non-relational part (the endowment value), which is a function of delay and nominal value of the component outcomes and a relational part (the contrast value), which is a function of the signed value difference between the outcomes, their interval and domain relatedness (i.e. whether or not the outcomes share the same domain). Delay and interval influence the endowment and the contrast respectively. Empirical investigations provide evidence for the contrasts model. Decision makers are capable of distinguishing between influences of delay and interval even when the two coincide and exert conflicting influences. Experiments using both money and non-monetary outcomes also show that preferences for improvement can be made more pronounced by shortening intervals and/or enhancing relatedness between the outcomes.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 0 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.