The purpose of the project was to determine whether offspring gender has an effect on parental favoritism. The research conducted prior to the experiment was grounded in sociobiological theory, meaning that we found that there are both social and biological reasons to favor one gender in offspring over the other. It was our hypothesis that parents would favor male offspring over female offspring. This hypothesis accords with both biological and social theory. Biologically it’s sound for animals to have male offspring because they can proliferate more often than females, and thus can carry on the family’s genes better. Socially it’s sound to prefer a male to a female, as males can “carry on the family name” and generally have more opportunities for success.
To test our hypothesis we performed a content analysis on articles from two magazines that focused on parenting. We tallied the number of times female children were referred to, the number of times male children were referred to, and the number of times the references to the children were gender non-specific. We also recorded whether the references themselves were positive, negative, or neutral. We found that male children were referred to more often than female children in the articles, and that non-gendered terms were used more often than either male or female terms. We also found that generally males were referred to positively more often than females, and that females were referred to negatively more often.
Although our data confirmed our hypothesis, we question the validity of the experiment and therefore the results. Based on the manifest content analysis of the parenting magazines our study was invalid, but in looking at the latent content of the magazines we can infer (but not “prove”) that male offspring are more favored than female offspring.
E.O. Wilson coined the term “sociobiology” in 1976 in order to describe a new paradigm of thought combining sociological views of human behavior and biological explanations. His new way of thinking elicited much criticism from his colleagues, who felt that his new paradigm could be used as scientific fuel for racism, sexism, classism, etc. Although his theory still causes much debate within the scientific world, it is an important paradigm when discussing questions of questions of “nature vs. nurture.” Although we will not be answering that question within our project, we will be pursuing questions concerning kin selection and gender status from within these views.
The purpose of the proposed project, then, is to determine the effect of offspring gender on parental favoritism. We will focus on the subject from the sociobiological paradigm. However, we will draw considerably from a number of other paradigms including feminist, evolutionary, and sociological. We have formulated a hypothesis that parents will tend to favor development of male offspring more than that of female offspring. We also suspect that this favoritism towards male offspring could be primarily a result of human social structures.
Although Wright believed that the status of the offspring’s gender depended on both biology and the socioeconomic status of the family, we will not be taking into consideration the later. Wright’s claim was based on the fact that women from lower-income families can “marry up”, that is, attract a male of higher economic status by way of beauty or servile behavior, depending on the society. However in today’s western society, families do not often benefit financially from the marriage of their offspring. Therefore, we feel this argument is invalid when applied to the society that we are studying.
Many studies have been done in the past to determine whether male or female offspring were favored. A study conducted by Adams and Lockard in 1980 attempted to measure the most “stable” biological unit of parent-offspring relationships by observing the frequency of parent/offspring combinations at the mall. They concluded that mother-daughter relationships were the most stable, as they were the most frequent. However, this study reflects only the gender boundaries that are accepted in this society, not the most stable relationships. For example, it is more acceptable for mothers to be seen caring for their offspring than males, and more acceptable for women in general to be a the mall. Therefore this study was not a good reflection of biological norms.
A study by Mackey (1976) took a very similar angle in attempting to discern which parent (male or female) spent more time with the children in public. This study encompassed several social spaces and nearly 20,000 people observed. However, Mackey’s study ran into several snags such as having to guess the relationship between the adult and child (it very well could have been the uncle, aunt, older brother, etc. with the child) and some of the aforementioned social stigmas that Lockard and Adams neglected to take into consideration.
We had originally planned to conduct public observations of our own, thinking we might be able to correct some of these problems through noting these researchers’ mistakes. However, we discerned that we would be recreating the same fallacies, as it would be impossible to differnetiate between soccialy prescribed gender boundaries and biological behavior. Speaking to those we would be observing could help solve some of the problems (what relation, personal views, etc.), but this would create the problem of the subjects knowing that they were being studied.
Through our research, we intended to carry these methods one step further and to not just simply observe that the relationships were present, but also observe the interactions between the biological units. However, this would present us with the same problem as the last study—it would reflect cultural norms regarding gender more than anything else. With the surveys, we would also run into the same problem. We therefore decided to conduct a content analysis for the following reasons:
Subjects are unaware they are being studied
It would decrease the amount that gender constructs would interfere with the data
it is cheap (accurate surveying can be expensive)
Originally, survey and observation were going to be the means through which the hypotheses were tested. However, it was decided that through survey and observation we would in fact not be measuring what we intended to measure. Survey questions such as “ How late do you allow your child to stay out?” would not determine how well the child was liked based on gender. Questions such as these would merely reflect ingrained cultural values regarding the parental behavior towards either males or females. This would make the results invalid and therefore useless to use in proving or disproving our hypothesis. Also, much of our data would have been useless because it would be equally easy to use the data to disprove our theory as to prove it. One could easily argue that protectiveness of female children represents favoritism. However, it is also possible to argue that permissiveness regarding females shows favoritism.
Since our originally proposed model was so flawed, we chose instead to conduct a content analysis on articles in popular parenting magazines to determine which gender was referred to more often, and thus theoretically more important. We began our content analysis by searching Miami University’s periodicals for magazines that dealt exclusively with parenting. Because there is no list of such magazines, we generated our sample by entering into the computer search system such obvious signifiers as “teething” and “toys”. We then observed the frequency of the magazines that contained articles on the topics. If the frequency was high, the magazine was either included or not included due to our previous knowledge of the magazine. It was for this reason that magazines such as “Family Circle” or “Ladies Home Journal” were not included in the sample, as they are not magazines solely devoted to parenting. Another determining factor was whether or not Miami University actually had the magazine. Due to these limitations two magazines were left to sample further from—”Parents” and “Parenting.” These were the only magazines that dealt with parenting exclusively and were available at the Miami University library.
We then counted the actual volumes of the magazines, and as there were forty seven volumes for the “Parents” magazine and thirty-nine for the “Parenting” magazine it was decided that eight volumes pulled from “Parents” and seven volumes from “Parenting” would be a good sample size. Random numbers were generated by dropping a pencil on a sheet of random numbers and beginning with the number the mark was made on. We then ended up with the numbers (34, 17, 12, 47, 25, 4, 1, 16) for “Parents” and (5, 30, 19, 7, 33, 22, 8) for “Parenting”. The corresponding volumes (beginning with the first according to year and month) were then pulled from shelves. Because the number of actual magazines within the volume varied so greatly it was decided that only one magazine would be sampled from each of the volumes.
The magazine chosen was also determined with the pencil drop test, with the only available numbers being the number of magazines that were in the volume. One article was then chosen from each of the sampled magazines. That article was the first article listed under the Features section, or in the case of the more recent “Parents” magazines the first article listed under the Family Life section. That article was then read for masculine signifiers regarding children (he, his, him, etc.), feminine signifiers (her, she, hers, etc.), and gender neutral terms (child, children, they, etc.) We also noted whether the term was used neutrally, positively, or negatively. A positive or negative value was assigned only in cases where the positive or negative connotation was extremely obvious. For instance, the only instance in which we observed “she” being used as an all inclusive term was when the magazine was discussing what to do with a child who threw “hissy fits”. This of course was not a perfect measurement, as male gendered pronouns are often used all inclusively, so it was difficult in determining whether the positive and negative correlations concerning male gendered pronouns were in fact valid.
Data Set #1 describes the extremely positive connotations we discerned from the texts. The x-axis represents the categories of gender we assigned (male, female, and gender non-specific) and the y-axis serves as the number of extremely positive terms found. There is only one bar rising from the x-axis showing a total of six such terms found when referring to males. We found zero obvious positive references for both the female and gender non-specific categories.
It was surprising to find that there were zero positive connotations for females in the 15 articles used. We had expected to find more positive allusions for males as compared to females, but we certainly did not expect a total absence for females. The cases recorded for the male category were only those that were completely obvious, such as the title of an article reading “How You Can Help Develop His Creativity” (our emphasis) when the article was referring to children in general. The gender non-specific category came out with zero as well. This category was slightly harder to interpret, being that the articles were taken from magazines devoted to parenting we expected the general tone of the articles to be positive towards children. However, we did not find any cases of blatant positivism to record.
Data Set #2 moves on to the extremely negative connotations we observed in our study. On this graph the x-axis once again represents assigned gender categories (male, female, gender non-specific), but this time the y-axis measures the amount of extremely negative connotations in the texts. There is only one bar representing data found in this set as well, except this time it is rising from the female grouping. A total of eight directly negative implications were found when referring to female children. Zero were found for both the male and gender non-specific categories.
Although we did not truly conceptualize “negative connotations” before we began, it was agreed that only blatantly obvious usage of the pronoun would constitute a negative connotation (the same guidelines were followed for conceptualizing “positive connotations”). We had expected to find few negative usages, as we were sampling from parenting magazines that would theoretically put a positive spin on childrearing. However, when there were blatantly negative connotations females were the gender being referred to (eight total as compared to zero for males and gender non-specific).
Data Set #3 describes the neutral terms (neither positive nor negative) found in the text. The categories of gender we ascribed are shown on the x-axis (male, female, and gender non-specific). The y-axis represents the number of neutral terms found in the texts. The numbers are much larger in this graph, as all three groups are being represented. Males totaled 226 neutral terms, females registered 112, and gender non-specific had 387.
This data set served mainly to give us an idea of how many times children were generally mentioned or referred to in texts under the categories of male, female, and gender non-specific. Our findings show that within the articles the generic connotation of “children” (or derivatives thereof) were the most commonly found with 387 total terms found. Males were referred to second most with 226 terms found. One should expect females to be somewhat equally represented as males within the texts, however we did not find that to be the case. Females only registered 112 terms, less than half as many as males. We attribute this shortage due to the male gender being more commonly used when speaking in generic terms of children. For example, one article about sleeplessness read “If your child’s sleeping problems persist take him to the family doctor or a specialist if necessary” (our emphasis). This literary folly certainly contributed to running up the numbers for the male category, but we found that more articles were generally talking about boys than girls.
A Chi-Square P-Value test was instituted to test our hypothesis. We entered the male and female results from all three data sets to compute a P-Value that would determine if our results were significant. A P-Value greater than .05 infers that the data has a high probability of randomness. That is, the data would have a chance greater than 50% of being random and without correlation to the hypothesis. This means the hypothesis would not be proven.
We computed a P-Value of <.0001, which revealed a startling significance of our data. The chances of our data being random were less than 1%, meaning the data we collected correlated with and essentially proved our hypothesis. There was a significant difference in the ways females and males were represented within the texts of the magazines used in our study.
Based on our data, it can be concluded that our hypothesis—males are more favored by parents than females—was supported and therefore true. However, as with any experiment, there are many flaws within ours that could have skewed the data and therefore made the experiment less valid. We had dropped our earlier experiments (survey, observation) because we felt that they would reflect cultural values rather than sociobiological behavior. However while conducting the experiment, we began to realize that our experiment was in fact doing the same thing. By measuring the gendered signifiers and the neutral signifiers in children’s magazines, we assumed that the three were mutually exclusive and exhaustive. However, often the words he, him, or his are used as neutral pronouns rather than gendered pronouns. Although it can be argued that that would also signify that males were favored over females, that would be delving into a completely complicated argument involving linguistics, history, and feminist writings. Therefore while it’s very interesting and relevant that there are generally more male gendered pronouns concerning children than female gendered pronouns, the information is not exactly valid to our experiment.
We also measured whether we believed the pronoun used was being referred to in a positive manner, a negative manner, or a neutral manner. However it should be noted that that we did not determine any set criteria concerning this variable before we began—it was completely subjective.
Therefore we both may have had different criteria for this variable. This measurement then was also not completely valid.
It comes as no surprise that our research was not exactly valid, as sociologists have been attempting to measure such qualities as “love” or “like” for decades, and have come up with equally as dubious measurements. Zick Rubin conceptualized the two ideas well in his article “How to Measure Romantic Love”, but the operationalizations that he came up with do not apply to parental love or liking. Even if they did, we would likely discover problems in applying those operationalizations to our research as it deals specifically with gender—and many attributes generally used to describe “love” or “like” would become twisted when applied to both sexes. For instance, we had originally decided to measure the amount of time a mother/father spent “gazing” or “looking” at her/his child, but this would only measure cultural beliefs concerning the independence and gender. Previous studies have shown that male babies are granted more independence than female correlates, and this would have to be taken into consideration in analyzing our data.
To this point our analysis has taken place within the realm of the manifest content of the magazines we sampled from. However, it may be useful to mention the latent content of the magazines (especially since we’ve determined that our data essentially proves nothing). As mentioned before, we attempted to measure whether the references were positive or negative. We decided to measure this after we had begun the experiment because we noticed that there were obvious instances in which the female pronouns were used negatively. (This is often referred to as inductive reasoning or grounded theory.) It was particularly noticeable because female pronouns were so rarely used compared to the neutral and male pronouns. Thus although we cannot “prove” anything from our data, we can infer from both the manifest and latent content of the magazines that males were preferred over females.
Adair-Gowaty, Patricia. Feminism and Evolutionary Biology. New York: Chapman and Hall, 1997.
Alexander, Richard D. Darwinism and Human Affairs. Seattle: University of Washington Press, 1979.
Brettell, Caroline B. and Carolyn F. Sargent. Gender in Cross-Cultural Perspective. London: Prentice-Hall, Inc, 1997.
Hammond, Dorothy and Alta Jablow. Women In Cultures of the World. Phillipenes: Cummings Publishing Company Inc., 1976.
Hinde, Robert A. Individuals, Relationships and Culture: Links Between Ethology and the Social Sciences. Great Britain: Cambridge University Press, 1987.
Lockard, Joan S. The Evolution of Human Social Behavior. New York: Thomand Books, 1980.
Lynn, David B. Daughters and Parents: Past Present and Future. Belmont: Wadsworth Inc, 1979.
van den Berghe, Pierre L. Human Family Systems: An Evolutionary View. New York: Thomond Books, 1979.
Wright, Richard. The Moral Animal. New York: Vintage Books, 1994.
Return to the Topic Menu
IMPORTANT: For each Research Response, make sure the title of the response is different than previous titles shown above!