Generational gender shift in names
Jessica Hom
Sam Kimbrel
Myly Nguyen
Patsy Partida
Anand Sampat
Linguistics 55AC
2007 December 10
Introduction
Our research topic came from an article published in the New York Times Magazine that discussed trends in name selection, entitled “What’s in a Name?” (Kean). It indicated that research had been done that shows that names are more likely to move from male to female usage over time than the other way around. Previous studies on the subject have been done by analyzing numbers from baby-naming books and the Social Security Administration’s database of names, and they show a definite trend for names to become more popular for female babies over time. (Barry and Harper) We felt that the next step would be to study people’s perceptions of names and to see if they changed over time by comparing the responses of younger and older people, particularly with respect to unisex names. A unisex name, as we use the term, is one that is given to both male and female people. Specifically, we set out to determine whether the shift in name usage over time is matched by a shift in the perceived gender of these names between generations.
Methodology
The most straightforward way of gathering data to research this idea was to use a survey. Other methods would not have provided the data we needed — it would be rather difficult to determine how people perceived a name through eavesdropping, and in-depth interviews would not have produced a large enough sample size. Our survey collected basic demographic information including gender identity, sexual orientation, hometown, ethnic identity, and age. The most important element of this was age, because we wanted to analyze the data with respect to generational shifts. The survey grouped responses for this into five-year categories to make analysis of the data slightly easier without losing too much detail. The main portion of the survey consisted of a list of twenty-one names that respondents were asked to place into one of four categories based on who they thought the names were appropriate for: male, female, unisex (but the respondent’s initial reaction was to think of a male), and unisex (initial reaction female). The names on the list were chosen to be a mixture of ones we felt to be exclusively male or female, as a form of control, and names that we felt were either unisex in today’s population or had been unisex in a previous generation. Many of them were taken from the list of names that Barry and Harper found as unisex names. (Barry & Harper, 231-232) The goal of this part of the survey was to get the initial reactions of people when they are presented with a name that is not directly connected to a specific person. In this way, we hoped to find a difference between generations in the perceived gender of names.
In addition, for each name they placed in the “male” or “unisex — initial reaction male” categories, respondents were asked to rate the masculinity of the name on a scale from one to five, and to rate the feminity of names in the other two categories. This was intended as a way to get finer-grained information on the respondents’ perceptions, and turned out to create a quite interesting result on its own (more on this later).
Surveys were distributed to members of the public, except in one case, where a member of the group traveled to the West Berkeley Senior Center to ensure that enough surveys were completed by older respondents. A total of one hundred and eleven surveys were completed, of which we used one hundred and nine to calculate results. The other two were insufficiently or incorrectly completed to the point of being unusable, indicating that at least two individuals had trouble understanding the directions given on the survey.
A copy of the survey is attached as an appendix to this report.
Method of analysis:
Our primary analysis of the data was done by separating the surveys by age group and tallying the number of people that placed each name in each category. We discovered that several age groups only had one or two responses, and so it was decided to group the surveys into four major age groups: respondents aged twenty-two and younger, respondents between the ages of twenty-three and thirty-seven, respondents between the ages of thirty-eight and fifty-seven, and respondents fifty-eight and older. These divisions were essentially arbitrary and made so that we could get a reasonably large number of completed surveys in each group; in retrospect, it might have been a better decision to pick ages that more clearly define generational boundaries.
Once the numbers were tallied up into these four groups, we calculated the percentage of each group that marked each name into each category and graphed these by category. These graphs are attached as an appendix to this report and show the percentage of respondents in each age group that placed each name in the given category. The masculinity/femininity ratings were averaged for each age group and organized by name and age group into a chart, which is also attached to the end of this report.
Quantitative results:
The first thing we realized was that we failed to take into account exactly how much data we would get from these surveys — using four categories of names and four age brackets creates sixteen numbers to be looked at for each name alone! Highlights of the analysis are presented here, because it would take an inordinate amount of room to discuss each name at length, and some of the names, specifically the ones we chose to be “controls” (e.g. Rebecca and John), do not demonstrate the trends we are interested in.
Two names that demonstrate interesting changes between age brackets are Ashley and Erin. Twenty-five percent of respondents 58 and older indicated that they perceived the name Erin to be a male name, while this response was only found in one-and-a-half percent of respondents 22 and younger, and not at all in the other two age groups. In addition, the name Erin was marked as unisex-male (that is, the respondent perceived the name as unisex, but the initial reaction was to think of a male person) by none of the 58-and-older group, by sixteen percent of the 38-57 group, nine percent of the 23-37 group, and seven percent of the 22-and-younger group. This is consistent with the idea of a generational shift from male to female in the perception of a name; however, it is a shift that is already well underway, as thirty-seven percent of the 58-and-older group placed Erin into the female-only category, and a different thirty-seven percent of this group indicated that they perceived Erin to be a unisex-female name. Similarly, Ashley was perceived as a male name by seventeen percent of the 58-and-older group, but none of the younger groups. This result can also be found for the names Leslie and Jordan, and to a lesser extent, Taylor and Morgan — it seems possible that perceptions of these names will remain unisex/gender-neutral rather than shifting completely to the female side of things.
A more interesting case is that of the name Andy — it may well be a name undergoing this shift in the current generation, as the following data show. Andy was placed into the male category by eighty-eight percent of respondents in the 58-and-older group, ninety-two percent of respondents aged 38 to 57, eighty-one percent of respondents aged 23 to 37, and only sixty-seven percent of respondents aged 22 and younger. However, the decrease in this response in the younger groups is matched almost entirely by responses placing Andy in the unisex male category! It seems likely that the name Andy is just beginning to undergo the shift toward unisex and then female perception, but this would remain to be seen conclusively by later studies.
As was intended by their placement on the survey, the “control” names Robert and John, and Sarah and Rebecca were rated almost exclusively as male and female, respectively. The single outlier is from a respondent who used only the unisex-male and unisex-female options, preferring to not put any names into the strict male and female categories.