Women's Studies 30

Survey of National Institutes of Health

Research Grants Awarded in FY95



We first intended to research different areas of United States Census Data, as was presented as an idea in an article we read for class on October 6th, 1998: Integrated Circuits / Segregated Labor: Women in Compter-Related Occupations and High-Tech Industries (see course syllabus).

As we investigated the possibilities of accessing this information through the Internet, we discovered that it was not readily available and it was not easy to sift through. We were having trouble narrowing down what exactly we would use for our topic, as well as deciding what parameters we would use in our investigation and how we would set up a control for the test. As we were exploring what was available on the web, we followed links and found ourselves looking at data on the National Institutes of Health (NIH) webpage.

We soon found a wealth of data available in the form of what grants the NIH awarded in the fiscal year 1995, in a collection of sites that are transcribed from the NIH publication called the Brown Books. This summary lists all of the grants awarded for that year.

The following text defines NIH grants and explains who is eligible for funding:

RESEARCH GRANTS - which make up the largest category of NIH research funding, are awarded to non-profit organizations and institutions, governments and their agencies, and to individuals and for-profit organizations for the support of research in health, medicine, and allied fields. Research grants provide funds for salaries, equipment, supplies, travel, and other allowable direct costs of the research as well as for indirect costs. Through the research grant mechanism, the NIH Institutes support research projects, centers, career development and other research and research-related programs. Small Business Innovation Research Grants (SBIR - Activities R43, R44) and Small Business Technology Transfer Grants (STTR - Activities R41, R42) are considered research grants, however, they are reported separately in the SBIR grouping.

We decided to focus on grants given to private researchers and their projects (not research centers), and compare both the numbers of male versus female researchers and the dollar amounts awarded to male researchers versus female researchers. We also distinguished between different types of cancer: male-specific (prostate), female-specific (breast, ovarian, endometrial, placental, cervical), and non-specific (non-gendered). More information about cancer, its causes, and possible victims can be found on the American Cancer Society webpage


The NIH grant data from the Brown Books is divided into six different sections for the fiscal year 1995 (FY95) data (index of Brown Books).

We randomly chose to include three books: numbers one, four, and five. Since the data for the Brown Books was listed alphabetically by grants awarded in each state, Book 1 included data from Alabama through Connecticut. Book 4 included data from the states of New York through Pennsylvania, and Book 5 spanned from Virginia to Wyoming and then included data for grants awarded to international research projects. The largest portions of our data came from states like California, Pennsylvania, and New York.

When judging what data would be included in our survey, we focused on certain keywords. The following is a list of the words that we considered to be indications that the research grant was for cancer research:

We also looked at the context of the research project's title, but we did not include all data that applied to a grant awarded to a group at a cancer research institute. For example, if there was a data point where a grant was given to a researcher at La Jolla Cancer Research Institute in California but we did not see any keywords or determine from the context of the title that the research was definitely involving cancer, we did not include the data in our survey.

Of a total of approximately 13,416 data points from the three Brown Books of NIH research grants from FY95, 697 were determined to be cancer-related. Within this group we divided the numbers into further subcategories:

We organized and analyzed all the data using Microsoft Excel 5.0. A summary table of data is available, or you can download all of the data we gathered:



There are many areas in which our data collection methods could be flawed.

The gender of the researcher who applied for the grant was determined by guessing from the first name of said researcher. There were many data points where we did not know the gender of a specific first name (and thus put the data into a category of 'unknown gender'), and most likely some names for which we guessed incorrectly. We placed all data from names that could be either gender (Jean, Pat) into the 'unknown' category. The knowledge of first names and their genders is also reflective of our personal knowledge of language and names. All data that was placed into the 'unknown' category was omitted from all analyses. A much more efficient way to determine the sexual breakdown of researchers' gender would be to obtain data from NIH statisticians.

There also exists the possibility that there were duplicate names, although we considered each grant to be individual unless the titles and researchers were exactly the same (in a few cases there would be a small initial grant and then a much larger one from a later date, seemingly the first grant for preliminary research and then further financial support for further research).

We tried to make our data sample as random as possible by picking to survey Brown Books that reflected a wide geographic spread of the United States and covered many different states. It would have been a more random sample had we gone through all of the Brown Books and looked at every third point (or some other arbitrarily picked number), which we would like to look at in the future.

We distinguished between data points that were cancer-related and those that were not by looking for certain keywords in the research project titles. These titles are not necessarily reflective of the actual number of research projects that were cancer-related - there were probably more studies that were erroneously omitted because of their titles, and some research projects that were not included because their titles did not reflect this.

We decided that breast cancer would be categorized as a 'female-specific' cancer, mostly based on the information provided by the ACS that seemed to be geared mostly towards women (see ACS breast cancer information page). It is possible for men to have breast cancer, but the occurrence is much smaller than breast cancer in women.

It would seem logical for men to receive larger grants from the NIH than women, as men are generally paid larger amounts than women even in the exact same field. The sciences have long been strongly male-dominated, so women are not as likely to participate in research such as we are surveying. In cases like this, since participation is already smaller for women than men, the added factor of lesser pay for women could cause a significant bias in the numbers of women versus the numbers of men in the field. An added possibility in this complex issue is the idea that men hold supervisory positions and women work as support or teams supervised by these men. In a case such as this, the men would be counted as the researchers but the women would not get credit for being involved in the research.

Not all of the research that we included in our survey was done on human beings - some involved laboratory work with rats or mice. This may or may not have any influence upon the findings of our survey.


Male-specific cancer research was not analyzed statistically because the n=2 sample size was too small.

The null and alternative hypotheses were the same for non-specific and female-specific cancers. The null hypothesis is that the mean research grant for male researchers is equivalent to the mean research grant for female researchers. The alternative hypothesis states that the mean research grant for male researchers is greater than that for female researchers.

We performed one-sided p-tests for both data sets. The p-value for non-specific cancer is 0.0005 at alpha = 0.05. The p-value for female-specific cancer is 0.0948 at the same alpha level. We reject the null hypothesis for non-specific cancer at this alpha level but can not do so for the female-specific data.

Non-specific Cancer: The boxplot comparing male and female researcher funding. Note that there are many more and many larger outliers for male researchers than for female researchers. (Created using DataDesk 5.0.1)

Female-specific Cancer: The boxplot comparing male and female researcher funding. The falilure to reject the null hypothesis is illustrated by the smaller number and smaller range of the outliers for male researchers' funding. (Created using DataDesk 5.0.1)

These analyses allow us to conclude that male and female researchers do not receive equal amounts of funding in non-specific cancer. Evidence was found to support the alternative hypothesis, which states that male researchers receive a higher dollar amount of funding from the NIH than female researchers. We could not reject the null hypothesis for female-specific cancer. This implies that male researchers receive larger mean grants than female researchers in this area of research.

In non-specific cancer, male researchers receive 83% of funding in contrast to 17% for female researchers. The numbers are slightly closer for female-specific cancer, where male researchers are awarded 64% of the total for this category of cancer and female researchers have 36%. In all types of cancer research (non-specific, male-specific, and female-specific) there were more male researchers than female researchers. For the specifics of these percentages, see the summary table.


From the statistical analysis of our data, we can conclude that male researchers obtain more funding from the NIH than female researchers in the area of non-specific cancer. Female researchers studying female-specific cancer receive 36% of the total dollar amount of NIH funding in this field, even though the number of female researchers in this field is much closer to the number of male researchers (41% female : 59% male).

We were unable to analyze male-specific cancer due to the small sample size of female researchers, so we cannot conclude that either the null hypothesis or the alternative hypothesis is valid. The split of funding does continue the trend put forth in the other two types of cancer research (94% of funding granted to male researchers, 6% granted to female researchers). This could be a reflection of the very large difference in numbers of male researchers in the field (93% of total) versus female researchers (7%). We are curious as to why female researchers are apparently so reluctant to be involved in research on male-specific cancer. Could this be a reflection of a bias against women in this field, or maybe a lack of interest on the part of women?

Women are less involved in cancer research in all three categories that we specified. We think that this could be reflective of the fact that scientific research is a heavily male-dominated field, and women might be discouraged from entering this field. There is also a strong cultural bias against women in many different careers, as men receive higher pay than women. This discrepancy might explain why male researchers are receiving higher-dollar grants than female researchers.

Name-recognition and reputation are also important criteria in awarding NIH grants, and if male researchers dominate the higher-up positions in research group ladders, then male researchers would appear to be awarded more grants. We would like to investigate the possibility that the research groups might be spearheaded by male researchers but the teams could be comprised of significantly more women than is reflected by the name of the grant applicant.

The large difference between the total dollar amounts and the total involvement in research of male-specific cancers versus female-specific cancers may be a direct reflection of the fact that we found many types of female-specific cancer but only one type of male-specific cancer. Because there are more possible areas for researchers to investigate in the category of female-specific cancer, there are more opportunities for grants to be awarded and more opportunities for researchers (of either gender) to become involved.

We regret that we were unable to find data from any year other than FY95, because we would like to perform a longitudinal study and assess the trends in differences between funding in the three categories of cancer, total cancer funding, and the involvement of male and female researchers.

Return to Women's Studies Homepage

Send message to Swarthmore College Women's Studies

e-mail us: annie and madhuri

last updated 12/8/98