quantile, trigonometric, random numbers and others. The reader should refer to the SAS
Institute Inc. (1993b) SAS Language Guide Version 6, for details.
The following SAS code will produce the exact probability for the specified binomial
distribution in the SAS log file:
data a;
p=probbnm1 (0.5, 17, 4);
put p=;
The function is used here in a data step, data a: defines a temporary data set in SAS given
the arbitrary name of ‘a’. p is the name of a variable which takes on the value of the SAS
function PROBBNML with three defined parameters, 0.5, 17 and 4. The first parameter
represents the population probability of success, the second parameter is the total number
of trials (sample size), and the third parameter is the number of successes (the smaller
frequency count in the two categories). The put statement tells SAS to write the variable
P (evaluation of the binomial function) to the LOG File.
Interpretation of Computer Output
The three lines of SAS code produce the following SAS output in the LOG File:
data a;
p=probbnml (0.5, 17, 4);
put p=;
run;
P=0.024520874
Note: The data set WORK.A has 1 observation and 1
variable.
Note: The DATA statement used 22.57 seconds.
The returned probability value of 0.025 is the probability that an observation from a
binomial distribution, with probability of success 0.5, and number of trials 17, has 4 or
fewer successes.
Worked Example (n> 25 )
Data from the study of upper limb injuries and handedness plasticity (Dellatolas, et al.,
1993) (used to illustrate Fisher’s exact probability test, in section 6.3) is used here to
investigate whether, for left-handed males, there is any difference between the probability
of young (≤6-years-old) and the probability of older (>6-years-old) subjects incurring
upper limb injuries. The sample data is:
Age when injuredLeft-handed males with upper limb injuries
≤6 yrs 7
6 yrs 59
100% 66
Statistical analysis for education and psychology researchers 176