American Chemical Society

1979 Comprehensive Salary and Employment Status Survey

L. Category which most closely approximates your present, or most recent, principal work function:
Research and development

Management administration of R&D

Ba resarch

Applied research development design
General management administrabon
(other than research development)
Marketing sales purchasing technical
Serres economic evaluation
Production quality cont

Foteniss analysis other lab analyses

Teaching teaching and research

Professor

Associate professor

Assistant professor

Instructor

Unranked

Writing editing abstracting libran, services

Data processing
Consulting

Other specify)

B.

C. "

D.

G. "

H."

L

M. Specialty which is most closely related to your present, or most recent, principal employment

The remaining questions refer to the MOST RECENT U S. PATENT covering an invention resulting from your work as an EMPLOYEE (whether for your present employer or a previous onei

U. Year this patent was issued: 19 ...

U.

V. Check all of the following statements that correctly describe the current status of this patent:
Patent has been assigned to employer

V.

ESTIMATING SAMPLING ERROR FOR PROPORTIONS

Upper and lower limits for the percents presented in this report may be estimated by using the table below. The table shows the approximate sampling errors for selected proportions and sample sizes. These sampling errors may be used to construct approximate "95% confidence intervals" for proportions. The sampling errors were computed, assuming the normal approximation to the binomial distribution, using the following formula:

In the table on page 10, for example, 6096 full-time employed respondents were classified as working in industry. The percent of this group who are inventors is listed as 40.5 percent (p=.405). A 95% confidence interval for this proportion may be approximated by taking n and p to be about 5000 and .40 respectively. The table shows an approximate sampling error of .014 (1.4%). Hence, the 95% confidence interval is (.405 - .014) to (.405 + .014) or .391 (39.1%) to .419 (41.9%)* If 100 similar estimates were made at this " level of confidence", about 95 of the true population proportions would be contained in their respective intervals. *Direct use of the formula gives .405.012.

FULL-TIME EMPLOYED RESPONDENTS:

INVENTORS AND NON-INVENTORS

by Employer (K), Degree (A), Sex (C), and Work Function (L)