selection, strain, or opportunity in grad student cheating?

aside from a few notorious cases where the accused quickly left a program, i don't hear much about cheating among sociology grad students.

a new self-report study finding high rates of cheating among business students is getting some press this week. business professor donald mccabe of rutgers, kenneth butterfield of washington state, and linda klebe trevino of pennsylvania state collected data from 5,331 graduate students at 32 colleges and universities in the united states and canada from 2002 to 2004. the researchers asked about 13 behaviors, including cheating on tests and exams, plagiarism, faking a bibliography or turning in someone else's work.

here is the prevalence of cheating by area of study:

Percentage of graduate students who acknowledged cheating in the past year:
Business: 56%
Engineering: 54%
Physical sciences: 50%
Med students/health care: 49%
Education: 48%
Law: 45%
Arts: 43%
Humanities/social sciences: 39%

i haven't seen the full study, but my guess is that the high prevalence rates are due to one or more common but less serious offenses. nevertheless, the study includes some good data on frequency and variety of cheating as well, which also seems to point the finger at business students.

the raw results raise some interesting questions. it is easy to formulate a selectivity hypothesis (e.g., business and law students are greedier), but there may also be more strain or competition in these fields (e.g., high-stakes exams and class ranks that matter) and, perhaps, more frequent opportunities to cheat (e.g., midterms and finals rather than seminar papers). is it selection, strain, or opportunity that places business students at the top of the list and social scientists at the bottom?