Concave utility, transaction costs, and risk in measuring discounting of delayed rewards.

Abstract

Research has consistently found that the decline in the present values of delayed rewards as delay increases is better fit by hyperbolic than by exponential delay-discounting functions. However, concave utility, transaction costs, and risk each could produce hyperbolic-looking data, even when the underlying discounting function is exponential. In Experiments 1 (N = 45) and 2 (N = 103), participants placed bids indicating their present values of real future monetary rewards in computer-based 2nd-price auctions. Both experiments suggest that utility is not sufficiently concave to account for the superior fit of hyperbolic functions. Experiment 2 provided no evidence that the effects of transaction costs and risk are large enough to account for the superior fit of hyperbolic functions.

Impatience and grades: Delay-discount rates correlate negatively with college GPA

“…In other experiments they reported an amount that they would be willing to receive at the time of the session in lieu of the delayed reward. Although in theory these procedural variations could lead to different responses, Kirby and Santiesteban (2002) found no differences in the discount rates that they generated.On each trial, the participant's reported present value of the delayed reward was entered into a sealed-bid, second-price auction. In some experiments participants bid against other participants (Kirby, 1997;1998), and in other experiments they bid against the computer (Kirby, 1998;Kirby & Santiesteban, 2002).…”

“…Although in theory these procedural variations could lead to different responses, Kirby and Santiesteban (2002) found no differences in the discount rates that they generated.…”

“…Thus, this model predicts that college students with higher discount rates should tend to place less weight on the future in current decisions, and should be more impulsive than those with lower rates. Consequently, we predict that discount rates and college GPA should be negatively correlated.To examine the relationship between GPA and discount rates, we obtained college GPA and SAT scores for students at two liberal arts colleges for whom discount rates had been estimated in previous studies (Kirby, 1997;1998;Kirby & Santiesteban, 2002 This phenomenon can be modeled using hyperbolic discount functions, which typically provide a better fit to empirical data than do exponential functions in which the discount rate itself is insensitive to delay. One hyperbolic function, which has been shown to fit animal and human discounting data quite well, is the following (Mazur, 1987):…”

An algorithm for identifying nonsystematic delay-discounting data.

“…Such measures have occasionally been used to quantify model fit in the delay-discounting literature (two instances are Kirby &Santiesteban, 2003, using RMSE, andMadden et al, 1997, using SSE). In addition, a variety of other approaches have been recommended as methods for evaluating nonlinear regression models based on model generalizability rather than goodness-of-fit, such as the Akaike information criterion (AIC) and the Bayesian information criterion (BIC) (Burnham & Anderson, 2002;Forster, 2000;Pitt & Myung, 2002;Zucchini, 2000).…”

Efeitos da Imprevisibilidade Familiar e das Diferenças em Função do Sexo Sobre a Propensão ao Risco, Exposição à Violência e o Desconto do Futuro de Jovens Universitários: Uma Abordagem Evolucionista

“…O escore de desconto do futuro foi obtido a partir das respostas individuais a nove questões sobre preferência por receber valores menores em um prazo mais curto ou maiores em um prazo mais longo (por exemplo, "Você prefere receber R$ 34,00 amanhã ou R$ 35,00 em 186 dias?"). O procedimento de cálculo do desconto foi baseado numa equação de função hiperbólica, denominado valor k (Kirby & Santiesteban, 2003, Wilson & Daly, 2004. Esse instrumento não apresenta normalização para respondentes brasileiros, mas tem sido utilizado em vários estudos mostrando adequação de construto (Ferreira, 2009;Ornellas, 2010;Howat-Rodrigues, 2010).…”