sensR (version 1.2.2)

discrimSS: Sensory discrimination sample size calculation

Description

Computes the sample size for any one of four methods: 2-AFC, 3-AFC, duotrio and triangle tests given the underlying sensory difference d, the type 1 test level and the required power.

Usage

discrimSS(delta, power, alpha = 0.05,
          method = c("duotrio", "threeAFC", "twoAFC", "triangle"),
          pd0 = 0, type = c("difference", "similarity"), start = 1)

Arguments

delta

the underlying sensory difference (larger than zero)

power

the wanted power (between zero and one)

alpha

the type 1 level of the test (between zero and one)

method

the discrimination protocol. Four allowed values: "twoAFC", "threeAFC", "duotrio", "triangle"

pd0

the proportion of discriminators in the population of interest

type

the type of test to be conducted

start

lower bound on the sample size. Specifying start at a value close to, but lower than the sample size will save computational time. If a too high value is given, this value is returned and the user can try a lower value for star

Value

• The sample size

Details

The sample size of the standard one-tailed difference test of "no difference" is obtained with pd0 = 0. The probability under the null hypothesis is given by pd0 + p0 * (1 - pd0) where p0 is the guessing probability defined by the method argument. The function uses one of the dedicated binomial families and the discrimPwr function

References

Brockhoff, P.B. and Christensen, R.H.B (2010). Thurstonian models for sensory discrimination tests as generalized linear models. Food Quality and Preference, 21, pp. 330-338.

See Also

triangle, twoAFC, threeAFC, duotrio, discrim, discrimSim, AnotA, discrimPwr, samediff, findcr

Examples

## Finding the necessary sample size:
discrimSS(1, 0.9, 0.05, "twoAFC")
discrimSS(1, 0.9, 0.05, "threeAFC")
discrimSS(1, 0.9, 0.05, "duotrio")
discrimSS(1, 0.9, 0.05, "triangle")

## A similarity example:
discrimSS(.2, .80, method = "triangle", pd0 = .2, type = "simil")