The Use Of Facebook And Twitter By Different Generations
For this written assignment, my research question focuses on the connection between media platforms and age. Specifically, I ask: “Does the use of Facebook or Twitter vary by generation?” The reason I believe this question is so important is directly tied to our nation’s current political climate. In the 2016 election, only “about half of the number of eligible millennials – those ages 35 and younger – voted” while over “45% of voters were above this 35-year cut-off.” This resulted in the successful election of a man who represents very old, antiquated, and frankly backwards views of what America represents and how it should be run. The lack of young peoples’ representation in this election resulted in Trump’s win, and, unfortunately, that time our nation has seen multiple government shutdowns, deportation acts, and assault rifle shootings.
The only way we can move forward from this current, deplorable state is by making sure young people’s voices are heard in the next election and this will happen if this group is persuaded to vote. There have been many studies that focus on millennials use of social media for networking or identity creation purposes, but my question can prove vital in figuring out how to get information to young adults about how to become politically active – specifically how to register to vote. I believe answering this particular question will be invaluable in the fight to change the current, undesirable political landscape.
In answer to my research question of: “Does use of Facebook or Twitter vary by generation?” I will say that I believe more millennials will use Twitter rather than Facebook. I believe this is the case partly because Twitter was the platform of my generation. The group that has been dubbed the generation that has “graduated college and had to enter a job market at its tightest since the Great Depression, is fiscally conservative, and socially liberal” also has come to “rely increasingly social media.” In other words, we are characterized as the generation “that’s lost faith in institutions” and resultantly put it “into the hands of [Twitter.]” In addition, I believe more millennials will use Twitter because Twitter champions free speech, and we are a generation that values this quality in the products we chose to interact with.
This is in stark contrast to what I believe are the usage patterns of the older generation. I believe older people will use Facebook for one central reason: it is a multi-use platform. With Facebook older people can connect with family members or even purchase home goods right on the page. Twitter has the sole purpose of social connection and information sharing; Facebook is flexible and can be utilized as a multi-purpose site, so I believe the older generation will prefer Facebook over Twitter.
Data and Variables:
For this assignment, I will be using an “extract” from the 2016 General Social Survey which is a public collection of social science information. Since 1972, the Social Survey’s target population has been English and Spanish speaking American adults. The sampling method that is employed during this survey is called a split ballot method in which “items are assigned to two of three ballots, each of which is answered by a random two-thirds of GSS samples.”
From this extract, I utilized two response variables and one explanatory variable. My first response variable is Facebook; my second is Twitter. For my first response variable of Facebook participants were asked “Which of the following social networking or social media sites are you a member or regular user of? Facebook?” Participants were able to answer “yes” or no,” and a total of 1,372 people were sampled using this question. As the person was tasked with choosing either “yes” or “no” we can class this first response variable as a nominal (or un-ordered) categorical variable.
As aforementioned, for my second response variable I chose Twitter. Questions associated with this variable were: “which of the following social networking or social media sites are you a member or regular user of? Twitter?” The participants could either answer “yes” or “no.” A total of 1,372 people were sampled for this variable. Like my first response variable, twitter is also a nominal (or un-ordered) categorical variable. As a result of the way both measures were taken for my response variables, I believe they are reliable but not valid. The split ballot method ensures the randomization of the sample so it is possible we could get the same number of people in each category again which makes the measures reliable. We cannot, however, observe these concepts directly, so we would have to use a criterion indicator in order to be able to say these response variables are valid.
My explanatory variable, age, was originally a quantitative variable as it numeric and represents a measurable quantity. In order to ensure my variable had variation however, I recoded age and made it categorical with “1” being millennials, or individuals 18 to 35 years of age, and “2” being individuals of greater than 35 years of age. For this specific variable, participants were asked their age, and they gave a number from “18-89 or older.” A total of 2,857 people were surveyed for this variable. In the variables original, quantitative form I would say that I believed the method of measurement made it valid, but not necessarily reliable. The question of one’s age ensures that the measurement captures the “true” concept of how old someone is that we are trying to encapsulate, so it would be valid. Unfortunately, as the standard deviation was 17. 7 I could not say for sure that the measure was reliable or precise. A high standard deviation indicates that there is high variability in the sample; and therefore, is not reliable. As I have now recoded the variable and made it categorical I would say it is now both valid and more reliable – which is in line with my reasoning for my categorical, response variables.
As Facebook is a nominal, categorical variable I chose to present information about it using a frequency table. You can see from the chart generated with “tab Facebook” that a majority of people use Facebook; exactly 1,022 people or 74.5% of people responded “yes” to using Facebook. Only 350 or a low of 26% of people responded “no” to using Facebook. For this nominal, categorical variable the most useful method for describing the central tendency would be the mode and since “yes” is the category that exhibits the most frequency, we can conclude that it is the mode for this variable.
For my second response variable of, Twitter, it again is most appropriate to use a frequency table. From my table (“tab twitter”) we can see that the majority of people don’t use Twitter as their preferred social media platform. Only 255 or a low of 19% of people responded “yes” to using Twitter while a high of 1,117 or 81% responded “no” or said that they do not use Twitter. Since “yes” is the highest repeated value – at a frequency of 1,117 – I can confidently say this is the mode for my data.
Finally, for my explanatory variable of age which I recoded as “agemil” I produced a frequency table. From my table I can see that the majority of people in the sample were in group “2” which is individuals 35 or older; 2,064 or 72% of respondents belonged to this category. A low of 793 or 28% of people were found in category “1” which is the millennial category. All three of my variables ended up being nominal, categorical variables and as such it does not make sense to include information on variability only the mode. This mode for this variable is category “2” – those 35 or older – with a frequency of 2,064.
Cross-tabulation Results:As all of my variables are categorical, I used cross tabulations to properly present the data on the variable’s associations. The first cross tabulation I generated, “tab agemil facebook, column chi,” shows that the usage of Facebook varies by generation. More specifically, individuals that belong in the 35 or older category of “2” have the highest percentage of Facebook use at 60.9% as compared to the low of 39.1% that millennials, or those in group “1” use. This is a large percentile difference (of almost 20%) between generations, so I can confidently say there is a difference in generational (or age group) use for this media platform with the elderly favoring Facebook over millennials.
My second cross tabulation, “tab agemil twitter, column chi,” shows almost the opposite result. A high of 54.5 percent of millennials use Twitter while only a low of 45.5 percent of the older generation use Twitter. This is almost a 10% difference with millennials favoring this specific social media platform. From these results, I can conclude that my two-part hypothesis – that millennials would prefer Twitter and the older generation would prefer Facebook – was correct and is supported by the data from this sample.
In order to generalize to the wider population, I selected a significance value of 0.05. Using the Chi-Squared statistic I generated in Stata I can see that the Pr, or probability level, for both “tab agemil Facebook” and “tab agemil twitter” is 0.00. Since these significance levels are less than my set significance level of 0.05, I would reject the null hypothesis of independence in favor of the alternative hypothesis (my hypothesis) that the variables are not independent of each other and I can conclude that these associations can be generalized to the wider population studied here of adult, English or Spanish speaking, Americans.
From my data analysis, I can see that my results are consistent with my hypothesis. I hypothesized that the use of Facebook and Twitter would vary by generation. Specifically, I believed that more millennials – those 35 and younger – would use Twitter and that more of the older generation would use Facebook. From my cross tabulations, I could see that not only did the respondents use vary by age but also that my hypothesis that more millennials will use Twitter was correct. Both the frequency percentage of “yes” varied within the columns across age groups and 55 % of millennials preferred Twitter while only 45% (or less than half) of the older generation preferred this same platform. I can now conclude that the calculations support my hypothesis and that these associations align with it.As this same trend of variation in frequency percentage happened within columns for my Facebook variable, I can also confidently say that there exists a generational difference in the use of these two social media platforms and as my Chi-Squared test showed this difference can likely be seen when applied to a larger population.