The development of categories and you will a good pilot reliability try survived 3 days, when day the new codebook was altered several times. To sample accuracy, a couple coders (scholar children) individually coded a batch from one hundred randomly chosen meme occasions (regarding 10% of one’s take to). The brand new contract speed is actually checked-out by using the Krippendorff ( 2004b ) attempt, with .66 just like the minimal rating to simply accept a varying once the required (Krippendorff, 2004a ). Eventually, one or two parameters, subculture (content) and photo list (form) gotten an insufficient score, and were taken off the fresh codebook, making forty-eight variables towards the real coding (Look for but in addition for the whole codebook of universal attributes from inside the all of our research). After reliability was actually situated, all of those other test try split among them coders.
Community data
We working community study techniques (Barabasi & Frangos, 2002 ; Newman, Barabasi, & Watts, 2006 ) so you can chart the content, participation, and you can form similarities between meme occasions as well as their cohesiveness. All of our stars regarding the network was indeed the latest meme occasions, in addition to their undirected connections were the typical posts, contribution, and you can means properties. Such as for instance, when the meme occasions A beneficial and you will B express around three simple attributes in the preferred, the wrap do weight about three. I developed three separate undirected networking sites getting blogs, function, and you may contribution, as well as a network that combines every issue along with her. The second is short for all round distance certainly one of meme instances within decide to try.
System analysis let pinpointing the quintessential main meme circumstances, simply put, individuals who common the highest level of articles, involvement, and means qualities with others, in addition to individuals who was basically a lot more peripheral, and therefore got novel or rare attributes ( ). Simultaneously, taking a look at the standard build of network, we are able to choose clusters regarding meme circumstances with high cohesiveness out of blogs, participation, and form. We are able to then consider which meme family had been a great deal more natural than just other people and you can tell you the web link ranging from common functions and members of the family cohesiveness.
Measuring the latest cohesiveness from meme families
Structural cohesion when you look at the companies features individuals measurements (White & Harary, 2001 ). We find the most simple one, and you can compared the latest ties certainly one of instances of the same members of the family having links outside of the members of the family. The newest cohesiveness varying was created to explore as to what the Lutheran dating sites free amount meme days was like other period from the same friends into the terms of general blogs, participation, and you will function. Very first, we calculated the brand new network knowledge (Barabasi & Frangos, 2002 ; Newman, Barabasi, & Watts, 2006 ) of each and every meme such as for instance, i.elizabeth., just how many common blogs, participation, and you may means characteristics so it shared with other days from the same family relations with hours off their family members. Following, each of one’s fifty meme families in our take to, i calculated this new imply number of common attributes its occasions mutual contained in this and you may outside the family. Eventually, the cohesiveness variable is defined as the real difference on the mean amount of prominent functions inside and exterior an excellent meme family members.
Measuring new individuality of meme families
The measurement out-of individuality height was designed to discuss just how unique an excellent meme relatives was a student in terms of the generic posts, involvement, and you may function properties. This basically means, individuality top shown the brand new pit between the common features of an effective meme family in addition to conventional (the best universal features in the entire decide to try). Basic, we computed the newest part of blogs, contribution, and you may form functions from inside the for each family; 2nd, we computed the portion of for each and every feature on entire sample; finally, for each meme family, we computed new square root of your amount of this new square differences between the two (like in new algorithm having basic deviation). It was standard to may include 0 and you can step one, where 1 represents best individuality from a good meme family inside regards to its universal functions.
