Автор: The development of categories and you will good pilot reliability try survived 3 weeks, during which day new codebook try changed a few times. To decide to try accuracy, a couple of coders (scholar students) individually coded a group of one hundred at random picked meme hours (regarding 10% of one’s try). The latest agreement speed is actually checked out utilising the Krippendorff ( 2004b ) attempt, having .66 while the minimum get to accept a changeable since necessary (Krippendorff, 2004a ). In the course of time, two details, subculture (content) and you will picture list (form) received an inadequate score, and was in fact taken out of brand new codebook, making 48 details into the real programming (Select but in addition for the entire codebook of your common features inside the investigation). After precision had been founded, other shot is actually split between the two coders.
Network data
We functioning community study procedure (Barabasi & Frangos, 2002 ; Newman, Barabasi, & Watts, 2006 ) so you can map the content, involvement, and you will setting similarities ranging from meme days and their cohesiveness. All of our actors about system was in fact this new meme era, and their undirected links were the common articles, participation, and you may function qualities. Such as for example, in the event that meme era An excellent and you may B show around three general features in the preferred, the link do weight around three. I developed around three independent undirected networking sites to have articles, function, and contribution, also a network that mixes most of the issue with her. The latter represents the general proximity one of meme times within our test.
Community analysis permitted determining the absolute most main meme days, to phrase it differently, individuals who mutual the highest quantity of stuff, contribution, and form characteristics with individuals, together with individuals who have been a lot more peripheral, and therefore had novel otherwise unusual properties ( ) meilleures site de rencontre pansexuelles. In addition, taking a look at the general structure of your system, we could choose groups off meme instances with high cohesiveness away from posts, participation, and you will form. We can following view and therefore meme group was basically way more natural than anybody else and you will show the link between general qualities and you will members of the family cohesiveness.
Measuring brand new cohesiveness out of meme group
Architectural cohesion when you look at the communities has actually individuals measurements (White & Harary, 2001 ). I chose the most simple you to, and you will opposed the fresh connections one of cases of a comparable household members which have ties beyond your relatives. The fresh new cohesiveness varying was created to explore from what extent meme hours were the same as other instances regarding same members of the family when you look at the regards to universal articles, participation, and you will setting. Basic, we determined the new network training (Barabasi & Frangos, 2002 ; Newman, Barabasi, & Watts, 2006 ) of each meme instance, i.elizabeth., the amount of popular content, involvement, and you can form features which distributed to other circumstances regarding exact same nearest and dearest in accordance with instances off their group. Upcoming, for each of your 50 meme household inside our shot, i computed brand new suggest number of common attributes their hours shared within and outside the household members. Ultimately, the latest cohesiveness varying are defined as the difference in the indicate number of popular qualities within this and additional a good meme family unit members.
Calculating the individuality away from meme family
The new dimension of individuality height was designed to talk about exactly how novel a meme friends was at regards to the common content, involvement, and you will means characteristics. Put simply, uniqueness level shown the brand new pit within general options that come with a great meme friends as well as the main-stream (the most common generic characteristics in the entire decide to try). First, we calculated the fresh new percentage of articles, contribution, and mode characteristics within the for each and every friends; 2nd, we computed the portion of each characteristic on whole take to; lastly, each meme nearest and dearest, i computed the newest square-root of your sum of the fresh square differences when considering both (as in new algorithm to possess simple deviation). This is standard to start around 0 and step one, in which 1 is short for the very best uniqueness out of a good meme family within the regards to their simple properties.
