Friday, 19 June 2015

first investigation group work

My first hypothesis is that there will be more emotion shown on women’s twitter featuring longer sentences. I predict the character count will contrast to males who supposedly do not go in to detail with expressing emotions.
The methodology my group and I thought of was to choose five tweets from a random girl and boy who have recently been active on twitter; pick out every three texts to use for date in order to contrast the two genders. Due to it being a small pool of data, we’ll have to randomize the selecting of tweets. There may be males who seem more effeminate than others which will become a limitation on the data research; this will become a problem for our predictions of there being a contrast.
In our data we can conclude that our hypothesis is somewhat correct as women show more emotion on twitter than men. 16% of technology features are used by females, dominating the 14% used by males, the two twitter users we have chosen support our hypothesis. Girls use emojis which are ‘classed/stated’ as being a weaker way of communication. Due to the lack of data we have found, limitations occurred because we couldn’t say our hypothesis is 100% correct- this method of more emotion being shown on tweets by females may not apply to ever female who uses twitter.
We created a second hypothesis predicting that females are more emotional using emojis which will allow us to use the Tannen’s Pairings theory.

During our investigation we found out that the hypothesis we based our investigation on was incorrect and didn’t work as successful as we hoped due to the data we collected. The limitations of the investigation we had for the second hypothesis is that there isn’t enough data to help support it. It’s not as predictable as we hoped so having more tweets would have allowed us to create a better investigation.

1 comment:

  1. Good - don't worry about the hypothesis not being supported: this isn't a limitation. You can equally set out to disprove a hypothesis or just explore if someone's research is still valid/valid for a different data pool. Be careful with subjective terms like 'showing emotion' and try and base what you are testing on theory and specific techniques that link convincingly to theories/concepts.

    ReplyDelete