Gender Differences in Technology:
Text Messaging Versus Talking on Cell Phones
By Rick Apple
& Joshua Johnson
Introduction: The Digital Divide
There is inherently a tendency in society for certain people to value certain tasks over others. The same can be said about people having different aptitudes for different things. This is to be expected, and, in fact, this must be the case when people themselves are not the same. A large portion of academic research in the social sciences is aimed at exploring and explaining these relationships. Particularly interesting is the case in which these different interests and aptitudes coincide with other divisions such as geographic distribution, wealth, or background. Correlating these things can give us insight into the nature of our psychology and society; although correlation does not imply and cause and effect relationship. We have learned in class that computer usage and gender seem to be related in a correlative way.
As we have learned, males tend to be generally more interested in using computer technology, and they enjoy different aspects of using computers than do girls. The following table showcases some statistics of males preferring computers relative to females:[1]
Computer camps / Video arcades / High school electives / Presence in computer areas / Computer Science majors / Computer science PhDs / AP Computer Science11:1 / 10:1 / 3:1 / 5:1 / 3:1 / 5:1 / 5:1
Clearly males outnumber females in a number of metrics pertaining to computer interest. We have also discussed ways in which the two genders prefer to use computers. Males tend to enjoy competitive games, sports and war metaphors, eye-hand coordination, and stimulating effects. Females, on the other hand, tend to dislike these things; they prefer programs with learning components, frequent feedback, and emphasis on words.[2] One explanation why boys tend to be more interested in computers is that much of the programming is geared toward these aspects that they enjoy.
Academic Studies
There have been a number of studies that support the findings mentioned above. One pertinent study to this topic was performed by Cooper, Hall, and Huff in 1990. They had noticed in previous studies that boys enjoyed computers more than girls, and they had also noticed that the majority of games seemed to be based on boys’ preferences. They wondered if the two observations were related. To test this, they introduced fifth grade male and female students to two games created to help students learn division. One was a male-oriented game called Demolition Division, and the other, Arithmetic Classroom, corresponded more closely with female preferences for computer games. They found that for Demolition Division, girls experience more anxiety, whereas for Arithmetic Classroom, boys were more anxious.[3] Therefore, the content of games may be partially responsible for the digital divide.
Another study, conducted in the same year, explains another cause of anxiety when using technology. Robinson-Stavely and Cooper tested Princeton University students at their aptitude with the game Zork, an adventure game. They also randomly included or did not include another person to be in the room while the test was going on. They found that the presence of another person in the room severely hampered the play of the female participants.[4] Perhaps the fact that females are less comfortable with people around when they are using computers further contributes to the gender gap between men and women in the classroom.
Introduction to Our Study
We wanted to conduct a study to gain further insight into the ways in which the digital divide is made manifest in technologies that people use on a regular basis, and particularly in our own context here at Princeton. We identified two different machines that Princeton students use frequently across both genders, computers and celluar phones, and decided to focus on cell phones. Our reasons were that threefold: the prevalence of cell phones is a newer phenomenon then is the prevalence of computers, it is possible to get by as a student without an in-depth knowledge of a cell phone while the same thing cannot be said regarding computers, and we thought that cell phones offered a few different features that made it easier to design a survey intended to capture gender differences in usage. Furthermore, cell phone use is becoming more and more prevalent on our campus, in the job world, and at home. If there do exist significant differences in the ways the two genders respond to cell phones, it would be advantageous to discover them sooner rather than later. In this way, we could gain better understanding as to how the increased prominence of cell phones will likely impact our culture, and take steps to eliminate or contain the most harmful effects of the gender divide in this area.
Our study was fairly simple. It consisted of a few brief questions which gave the respondent the opportunity to report on their use of cell phones for various features. Originally, we intended to just measure the comparative differences between talking and text messaging, thinking that boys would use text messaging comparatively more since text messaging involves a more comprehensive understanding of the technology and requires more “button pressing”. These button-pressing activities, we thought, would be akin to using a computer, and therefore males would prefer them. As we developed the survey, we decided to include additional questions to measure differences in two more features, gaming and ringtones. Some cell phones come equipped with built in video games, and we were guessing that boys would use these features comparatively more. Ringtones are additional ditties that one can download off the internet to replace the standard and often shrill ringing sounds which cell phones come equipped with. Our survey was designed so that people could complete it very quickly (on the order of a minute) so that we could gather a lot of data in the limited amount of time we had for the project. We suspected that unless we had a large amount of data, we would not be able to isolate the particular effects we were searching for, or that we would not be able to show that our results were statistically significant.
We sent the following survey out to between 200 and 300 Princeton University students:
Do you own a cell phone (Y/N)?
Have you downloaded additional ringtones (Y/N)?
If you do own a cell phone, please answer the following three questions
regarding your cell phone usage. For each question, answer 1-5, where
the numbers mean:
1. Multiple times daily
2. About every day
3. Every week
4. Every month
5. Almost never
How frequently do you use your cell phone to make phone calls?
How frequently do you use your cell phone to text message?
How frequently do you use your cell phone's gaming features?
The first question was simply a weed-out question to eliminate those individuals who do actually own cell phones. The second question provides a simple yes or no answer to whether the person has gone through the process of downloading additional ringtones. And the final three questions are designed to gauge how frequently the person uses the various functions mentioned above: talking, text messaging, and playing games. The respondents answered these three questions by designating a number according to the coding scheme also contained in the email. We separated on our own which responses where from males and which were from females as we manually compiled the data into an Excel spreadsheet.
The survey was thus designed to reflect any actual differences in how males and females use celluar phones. Our central hypothesis was that boys would use the more involved and complex functions of the cell phone more than girls (gaming, text messaging, and downloading more ringtones). Because these functions require higher level of comfort with the technology and involve much more “button pressing” we thought that boys would likely use them more, as they do computers. Although we did track how frequently the two genders use the phone’s primary talking function, this was not the central point of our study and the results are included as Appendix 2. We thought that even if males and females generally use cell phones for talking roughly the same amount, that males would be more likely to use the more in-depth technological functions. We especially thought that this would be true regarding the games, since those involve competition and it is a common fact that boys play video games more than girls do. On the other hand, we were not as confident regarding the text messaging since it involves words and communication, two things which are known to appeal strongly to girls.
For an appropriate statistical test to measure the significance of the observed gender differences in our survey response, we employed the following test statistic[5]:
The Z value is equal to the observed difference in the two sample population means, divided by the square root of the sum of the sample variances over the sample sizes, where M is the number of boys surveyed and F is the number of girls. Because of our large sample size, the Central Limit Theorem guarantees that (μB – μG ) has an approximately normal distribution regardless of the underlying population distributions. And furthermore, using the sample variances σ2B and σ2G ensures that our Z statistic has approximately a standard normal distribution. Using this test, our null hypothesis is that there is no significant difference between the two observed sample means for the two genders, and our alternative hypothesis is that one of the means is higher than the other:
As is typical in many statistical exercises, we elected to use a two-tailed confidence interval of 95%. Therefore, when the Z statistic is between -1.96 and +1.96 we fail to reject the null hypothesis (or stick with the null hypothesis), and when the Z statistic is outside of this range we reject the null hypothesis in favor of the alternative hypothesis.
If the Z score falls outside of the 95% confidence interval, than that means that there is less than a 5% chance that we would have obtained the results we did obtain, if there were no differences between the genders. The actual likelihood of obtaining a given result is indicated by the P value, to which the Z scores correspond. We will conduct this statistical test on each of our variables to test for gender differences in use of particular cell phone features, and report the relevant Z score and P value, along with the corresponding conclusions indicated by the statistics.
Results
The graph above depicts the cumulative results of our survey. We collected 161 responses from Princeton students who did have cell phones, the other responses were set aside because they did not impact our study. Conveniently, we received a roughly equal number of responses from boys (82) and girls (79). Above, is shown the average response. While the last three categories were scored on the five point scale as explained in the survey itself, the first question “download”, refers to the proportion of students who have downloaded additional ringtones and is scored out of 1. A more comprehensive tally of the breakdown of responses is available in Appendix 1. Since higher numbers on the survey response indicate that the individual is using the particular feature less frequently, the data do show that, on average, men use all of the features more than women. The real question is whether these observed differences in the data are statistically significant, and whether they can be used to draw interesting conclusions regarding our original hypotheses.
Ringtones
Z score: 0.560
P value: 0.7123
Result: We did not find that boys are more likely to download ringtones
As shown on the graph, the data indicate that boys and girls were roughly equally likely to download additional ringtones, with 24 boys and 20 girls responding that they had done so. Applying our statistical test, we calculated a Z score well within the 95% confidence interval, and consequently we fail to reject the null hypothesis. This does not mean that boys and girls do in fact download ringtones the same amount, it only means that we cannot conclude otherwise based on our data.
Text Messaging
Z score: -2.764
P value: 0.0029
Result: We did find that boys are more likely to text message
According to the P value, there is a less than one percent chance that we would have obtained the results we did obtain if boys and girls actually used text messaging the same amount. Since the Z score falls outside of the 95% confidence interval [-1.96, +1.96] we reject the null hypothesis in favor of the alternative hypothesis and conclude that males are actually more likely to text message than females.
In computing the Z statistic we compared the two mean responses over the 5 possible options. The mean response for males was about 3.2 whereas the mean response for females was about 3.8. As explained above, even though these responses are not normally distributed, and indeed the bin size is inconsistent ranging from multiple times a day to never, the test statistic is still appropriate because of the large sample size. The graph of the responses is below.