Blog Post

Theoretical Justification

When analysing social media, more specifically, analysing social media usage in terms of acquiring news, age is a predominant factor to consider when looking at information. Social media and age are commonly linked variables, as different age groups use social media differently. In saying that, outsourcing news from a social media platform such as YouTube, where the platform focus is not necessarily driven to provide news to the public, but rather as a form of entertainment and general media. This is an important factor to consider when looking at age and social media, as this shows more of the dependency younger users have for social media, versus older users. Traditionally, news is sourced from conventional sites, such as online news channels, journals, etc., but in some user’s case, news can be sourced from platforms other than the conventional. This specific analysis of Youtube usage for news and age, contributes to the overall research of social media by uncovering the dependency and variety platforms provide, and how each age uses those resources.

To begin, in a journal written by, Özgüven, Nihan; Mucan, Burcu, titled, ‘The Relationship Between Personality Traits and Social Media’, the authors discuss the different personality traits and connections that link to social media usage, one being age. The authors write,

“Social media users are most commonly young adults (aged 18 to 31); three-quarters of adult Internet users under age 25 have a profile on a social media site (Lenhart, 2009). Social media websites are designed to be widely accessible and initially attract homogeneous populations, so it is not uncommon to find groups using websites to segregate themselves according to nationality, age, educational level, or other factors that typically segment society, even if that was not the intention of the designers (Boyd & Ellison, 2008).” (Özgüven, Nihan; Mucan, Burcu, 2013).

The authors distinguish that although the intended use for the platform may be one thing, the users ultimately decide the desired use. This can be applied to younger users and the link to Youtube for news. Although Youtube may have intended the audience to use the platform one way, the audience ultimately adapted based on their own needs and preferences.

Youtube has also become a staple for many when it comes to reviews, information and so on. The content creators have even adapted to create channels providing current stories, news, ranging from global news to less popular stories. Anyone seeking news information while using the platform can also search previous news stories from any channel on Youtube. The platform has become a place for visual and auditory database.

In accordance to Youtube as a popular entertainment source, there is a mass market for political usage and political campaign related news. In a journal written by, Albert L. May, titled, ‘Who Tube? How Youtube’s News and Politics Space Is Going Mainstream’, the author researches how Youtube has adapted and managed to become a predominant source of news. The author writes; “Based on data collected through TubeMogul, this study suggests that corporate media as a group, driven largely by AP, have steadily increased their share of the YouTube audience, encouraged by YouTube’s pursuit of partnerships with established news organizations.” (May, 2010). Youtube’s connections with major news organizations allows for audiences of any age to tune in at any time and in any location. News sourced for Youtube rather than cable opens the possibility of viewers as there is more convenience and promotes trending videos to the users.

In summary, Youtube has increasingly developed their online presence in regards to news. The platform not only partners with existing channels and organizations for viewers, but also general content creators on the site produce less traditional channels that also dominate the view count, catering to news stories and relevant topics. From this information, it is important to further breakdown the association with age and this news consumption style, as the platform and the users are in constant shift, becoming more and more codependent on one another.

Variable description

Independent Variable: That use of Youtube as a source of news.

Dependent Variable: Age.

The level of measurement of our independent variable is nominal dichotomous, as you either use Youtube for news or you don’t, and the level of measurement of our dependent variable is ratio, as age is a continuous variable. The Independent variable question of Users that use Youtube for news is asked as, “Which,if any, of the following have you used for finding, reading, watching, sharing or discussing news in the last week?” The question is then broken in sub categories, one category specifically asks if users use Youtube for news. We make this more specific by assuming that anyone that uses Youtube for news has an active Youtube account.

The Dependent variable question of age was labeled as “what is your age in years?”

Hypothesis

H0: The mean age of people who have not used Youtube for news is equal to the mean age of people who have used Youtube for news.

H1: The mean age of people who have not used Youtube for news is not equal to the mean age of people who have used Youtube for news.

Test description:

In order to accurately examine the relationship between our two selected variables, we had to employ an independent T-test. The main purpose of this test is to determine the significance of scale differences (interval/ratio) between categories of nominal (dichotomous) variables and it uses the mean values of each group as a unit of analysis. In simpler terms, the T-test examines the implication of mean differences between nominal dichotomous groups e.g. the age . Mean differences that are statistically significant imply that the established relationship between the IV and DV is supported enough to be accepted or rather, there is a low probability that this relationship occured by chance. One the other hand, mean differences that are statistically insignificant imply that the established relationship between the IV and DV does not have enough support to be accepted or rather, there isn’t a low enough probability that this relationship occured by chance. One final important thing to remember is like any other hypothesis testing instrument, the T-test comes with certain conditions/assumptions that must be respected. This is what we will cover in the next section.

Test Assumptions:

Some of the assumptions of the T-test we already covered in the previous sections: the IV must be a nominal dichotomous variable and the DV must be a scale variable which is to say either an interval level variable or a ratio level variable. These two assumptions as we established in the first section were met.The remaining assumptions of the T-test are as follows:

• DV should be normally distributed in IV — no high skewness or kurtosis
• Populations for all groups in the nominal variable should have equal variances
• Random sample

In their methodology, the researchers in charge of this survey noted that their sample was created through random processes, so we know that our last assumption was also met. We shall discuss the remaining assumptions in depth in later sections. For now it is sufficient to note that all the assumptions were met.

Table Description:

The first table that is represented below provides us with general information about the variables we are interested in, ( age and the use of youtube as a source of news). The most important values to look on this table are the N(number ) value, mean value and the standard deviation. The N value reveals the portion of the population that each group represents. The mean value is mainly useful when analysing scale variables (interval/ratio). In our scenario it represent the average age of members of each group. The standard deviation represents the degree to which the age of members of each group departed from the mean age. This is important because it allows us to know how spread out the ages of members of each group are relative to the average.

The second table, we have represented here is the independent sample T-test. As previously mentioned, this table’s main purpose is to establish the significance of the mean difference between nominal dichotomous variables. The first value to pay attention to when looking at this table is the Levene’s Test for Equality of Variances. As the name suggests, it looks at the level of homogeneity/similarity in the distribution of groups. This is one of the assumptions stated in the previous sections. As you might note there are two columns in this table, the first titled “Equal variance assumed” and the second “ Equal variance not assumed”. In order to know which of column to look at when discerning the T-test result , we must look at the “sig” value of the first column. If this value is bigger than .05 (P>.05), then we cannot confidently assume variance, which of course means our 4th assumption was violated and so we must use the second column to extract our T-test result. Alternatively, if this same value is smaller than .05 (P<.05), we can confidently assume variance ( our 4th assumption is met), this means we must use the same column to extract our T-test result. The most important row to look at when determining the result of the T-test, is “sig 2 tailed”. This shows us the extent to which our results are likely due to chance. Similar to the levene’s test, the acceptable significance of difference in means is .05 or lower. This means, everytime we see this value as .05 or less, there is a statistically significant difference in means and thus we accept our alternative hypothesis. Alternatively if this value is bigger than .05, we can assume that the difference in means between groups is not statistically significant and thus we accept our null hypothesis.

The third item we used to analyse data was a histogram representing the frequency distribution of members of each group (in terms of age). These diagram shows us some measures of central tendency e.g mode. It also allows us to validate our 3rd assumption, which requires that all values of the dependent variable are normally distributed into the independent.

Results of Test:

Within the frequency distribution table is, we see that the mean age of people who do not use Youtube for news is slightly higher than that of people who use Youtube for news. The difference in means is, 4.26. This mean that on average, people who do not use youtube for news are 4.26 years older than people who use youtube for news.

Next represented is the independent sample T-test. Showing a null hypothesis as having the mean age of people who have not used Youtube for news as equal to the mean age of people who have used Youtube for news. Using the Levene’s test for equality of variances, we are able to accept the alternative hypothesis because the Sig. value is smaller than 0.05, meaning the variances are not equal. In order to double check this result we look at the std.deviation in the frequency distribution table. The numbers represented have a significant difference between them, making it a good assumption that the variances are not equal.

In relation to the histogram we found that the distributions of both groups had normal tendencies, which validate our 3rd assumptions. This said however, there are limitations to this which will be discussed.

Implications:

The implications of the the results of our testing concludes with the final hypothesis that; The mean age of people who have not used Youtube for news is not equal to the mean age of people who have used Youtube for news. In other words our final results showed that the people who are using youtube to generate their news coverage are of a younger demographic. Younger demographics in our current day society such as Millenials are a generation who is much more connected to the outside world through their unlimited access to technology. Relying mainly on social media platforms such as; Facebook, Instagram, and Youtube for their day to day news updates. In comparison the Baby Boomer generation who is much more traditionalist in their news coverage turns to radio and print coverage to attain their news coverage. This fact is proven within our results showing that people who are getting their news coverage from youtube are on average 4.26 years younger then those who are getting their news from other sources.

Limitations:

As with every study, limitations were found. In this study of age and Youtube usage for news, the limitations includes test limitations and data or sampling limitations. The dataset used to compare the specific variables is the ‘Reuters International Digital News Report’. From this data set, the sampling method included Canadian citizens 18 year of age and older, meaning younger age groups are lost when comparing the variables of age and Youtube use for news. As explained in previous sections, age has a strong correlation with Youtube usage for news, specifically younger age groups. By limiting the data to 18 years and older, many younger age groups are not being represented in the data, that would ultimately produce results that do not reflect the real canadian population.

In addition, although four of the assumptions were fully met in the data, the histogram (figure 6) is displaying problems with the distribution of the data, and depicts the third assumption was only partially met. In the ‘no’ group in Figure 6, the distribution, or the height and way the bars are grouped together, depicts a normal distribution but can also be classified as bi-modal, meaning it has two peak points. The two peaks in the ‘no’ group of figure 6 means that there are two groups of age ranges that frequently do not use Youtube. When looking at the ‘yes’ group in figure is, there is a display of negative kurtosis, which implicates that the third assumption was not fully met. Lastly, the negative kurtosis (Platykurtic) from figure 6 is only partially met as it is not a normal distribution.

Bibliography

May, A. L. (2010). Who Tube? How YouTube’s News and Politics Space Is Going Mainstream. The International Journal of Press/Politics, 15(4), 499–511. Retrieved from https://journals-scholarsportal-info.proxy.bib.uottawa.ca/pdf/19401612/v15i0004/499_wthynapsigm.xml

Newman N, Fletcher R, Kalogeropoulos A, Levy DAL and Nielsen RK, “Reuters Institute Digital News Report 2017,” http://www.digitalnewsreport.org/.

Özgüven, N., & Mucan, B. (2013). THE RELATIONSHIP BETWEEN PERSONALITY TRAITS AND SOCIAL MEDIA USE. Social Behavior and Personality, 41(3), 517–528. Retrieved from https://search-proquest-com.proxy.bib.uottawa.ca/docview/1504174048/fulltext/2817455C668E4032PQ/1?accountid=14701

(Figure 6)