Updated on Aug 25, 1999

Programs in BASIC   List in FORTRAN   JavaScript

You may be able to save these programs as text from your Browser, or Copy and Paste them into a text processor. They should work with most BASIC compilers, such as GWBASIC for PC and MS QuickBasic for Mac.

Except as noted below, all are copyright Michael H. Birnbaum, all rights reserved. They may be used for educational and scholarly purposes without charge. No warranty is expressed or implied. No guarantees can be made for the software.


DMCALC is a BASIC program to compute the values of risky gambles according to several theories of decision making. Includes the Configural weight, TAX model of Birnbaum & Chavez (1997) and Cumulative Prospect Model with a 2-parameter W(P) function. It handles up to 10 nonegative outcomes. The program is set up with the examples in Birnbaum's chapter (in press) in the Festschrift honoring Ward Edwards, edited by Shanteau, et al., to be published by Kluwer. An on-line version in JavaScript is available by clicking here.


DMCALC2 is a BASIC program to compute the values of risky gambles according to the configural weight, RAM model of Birnbaum & McIntosh (1996) and Cumulative Prospect Theory of Tversky & Kahneman (1992). Rob Bailey has written a simplified, on-line version of this program in JavaScript, which you can try by clicking here.


ADDFIT is a BASIC program to fit an additive model to a two-factor experiment. Seminar Students: You can use this program to check your work on Homework No. 5A (Mystery Table) and 5B (Additive Resistance). You can also use this program to fit a subtractive model, but you will have to reflect (i.e. reverse the signs) of the subtrahend (difference = minuend - subtrahend).


MULTFIT is a BASIC program to fit a multiplicative model to a two-factor experiment. Seminar Students: You can use this program to check your work on Homework No. 5A (Mystery Table 1, Birnbaum, 1973), Pizza Problem, and Salary problem (Birnbaum, 1983). You can also use this program to fit a ratio model, but you will have to invert (i.e., 1/s) scales for the divisor (ratio = dividend/divisor). This program uses an iterative method to get a least-squares solution.

List of Programs in FORTRAN   List in BASIC   JavaScript


RDFIT.FOR is a FORTRAN program to fit two arrays of data simultaneously. The data are judgments of "ratios" and "differences" of magnitude in the same factorial designs. The program can fit one operation theory and several versions of two operation theory. This program is described in Birnbaum (1980).

RDFIT.DAT is a file of data showing setup for RDFIT. Included are mean judgments from Birnbaum's (1978) Cognitive Theory chapter and mean judgments from Birnbaum & Sotoodeh (1991).


RECIPE.FOR is a FORTRAN program that gives least-squares estimates of weights and scale values in a relative, constant weight averaging model to judgments collected in the recipe design. In recipe, a factorial design of three informational factors, ABC, is intermixed with two-way designs of AB, AC, and BC, and single factor designs of A, B, and C, each created by leaving out one or two pieces of information.

STEPIT.FOR is a fortran subroutine that can be used to minimize or maximize a function of several parameters. It was written by Chandler (references included in the file). This subroutine must be appended to recipe to complete the program.

RECIPE.DAT is a sample data setup showing how to arrange data for recipe. The data are hypothetical.


RTFIT.FOR is a FORTRAN program to fit three arrays of data simultaneously. The data are "LESS" choice Response Times, "MORE choice response times, and judgments of "differences" of magnitude in the same (subset of) a factorial. The program fits the theory of Birnbaum & Jou (1990), which accounts for the distance effect, serial position effect, and semantic congruity effect as well as "difference" judgments using the same scale of subjective magnitude.

RTFIT.DAT is a file of data showing setup for RTFIT. Included are means from Experiment 1 of Birnbaum & Jou (1990), COGNITIVE PSYCHOLOGY, 22, 184-210.


MONANOVA.FOR is a FORTRAN program written by Kruskal and Carmone to transform data to additivity. References are included in the file. This version has a small modification to allow the program to create a file of the transformed data, which is useful for further analysis of the transformed scores.


MONANOVA.DAT is a file of data showing the setup for MONANOVA. Included are means from Birnbaum(1974; 1982) and from Birnbaum (1978).


TAXFIT.FOR and CPTFIT.FOR are FORTRAN programs that fit choices between gambles, according to Birnbaum and Chavez (1997) configural weight TAX model and Tversky and Kahneman's model of cumulative prospect theory (CPT). Data may be forced choice, or they may also include judgments of strengths of preference between gambles. Programs allow compromise between log likelihood criterion that only depends on choice and sum of squared deviations that depends on both choice and judgment of strength of preference. Variants of these programs were used by Birnbaum and Chavez (1997). Sample data from Birnbaum, Patton, & Lott (1998) are included here, TAXFIT.DAT and here: CPTFIT.DAT, to illustrate how to set up the data. You will also want to download the following RTF file, which can be read in most word processing programs. Handbook for TAXFIT and CPTFIT

List of Programs in JavaScript   List in BASIC   FORTRAN

surveyWiz and factorWiz

SurveyWiz and
factorWiz are Web pages that make Web pages that run simple surveys and experiments. SurveyWiz allows the user to enter up to 90 personality or agree-disagree survey items or items requiring either numeric or text response from the subject. FactorWiz creates the HTML Web page for a factorial design. You specify the number of rows and columns, and the values of the factors, and this Web page does the rest. Instructions for use are contained in the Web pages themselves.
ChoiceWiz with Tables with Colors for Dependent Gambles

taxcalculator and cwtcalculator

taxcalculator and cwtcalculator are JavaScript calculators. These work only with Netscape, and they calculate predictions for the TAX model, the RAM model, and the CPT model with parameters specified by the user. Taxcalculator calculates under the TAX model of Birnbaum and Chavez (1997), and cwtcalculator calculates under the RAM model of Birnbaum and McIntosh (1996). Taxcalculator uses the 2-parameter weighting function of Lattimore, et al and Tversky and Fox. cwtcalculator uses the 1-parameter weighting function for CPT given in Tversky & Kahneman (1992).

Return to Birnbaum's Home Page

nsf logoThis material is based upon work supported by the National Science Foundation under Grant No. SBR-9410572. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation.
Jan 14, 1999