I made a graph recording the number of words per text. This was done by looking at texts which had been sent during a 20 minute period. I counted all individual entities to be separate words, such as “k”, but for contractions, such as “can’t”, I counted that as two words. However, for the abbreviation “lol” I counted it as one word as it is used in a different way than the constituent words. Some of the texts were taken from a text conversation when the sender was totally sober and had not drunk anything, the other texts were taken from the same sender’s conversation but at a time when they had drunk a few glasses of wine and who was noticeably slightly inebriated.
As you can see from the graph without alcohol less texts in general were sent and each text tended to be shorter. You can also see that without alcohol there was a significant increase of very simple sentences. This would suggest that without alcohol the sender is happy to say things simply without additional detail or embellishment. I would think that this is because alcohol lowers inhibitions and so the additional information which people want to say but normally leave out is included in texts. This would also partially explain the increased number of texts as someone who wishes to convey lots of information is more likely to send another text to add further detail to a previous one.
You might also note that on average without alcohol the texts were four words long which is consistent with very basic sentences such as Subject Verb Object which also includes a few articles e.g. “I watch a film”. Whereas when the sender had alcohol the mean text length was six words long and tended to include more unneeded words such as the text “Anyway I can cook that”. But when you look at the graph you can see this is misleading as the more average text length is actually slightly higher but is skewed by the large amount of single word texts sixteen with alcohol and eleven without. If you find the mean ignoring the single word texts you get a mean of six and a half without alcohol and seven and a half with. This shows that, excluding the single word texts, the texts with alcohol is only a word more than without. This suggests that the alcohol only has a slight effect on the sentence length of texts. With single word texts however, which tend to be simple affirmatives, such as “k” and “yeah”, there is nearly a fifty percent increase in these texts when the sender has had alcohol.
I also counted the number of contractions and abbreviations and found that without alcohol the sender only used ten whereas with alcohol they used thirty eight. This is nearly a fourfold increase which is rather large. This would suggest that with alcohol the sender wants to use more words but is impatient and so used contractions to make long sentences with less typing.
From these results I would conclude that the smallish amount of alcohol has only a slight effect on the structure and frequency of texts, however the effect is still noticeable. This effect seems to be making the sender more loquacious as their texts were on average longer and they was a significant increase in the frequency of texts. The increase in contractions also suggest an impatience and desire to say more in the same amount of time. However, my conclusion is not very well supported as I have only analysed texts sent in a twenty minute time period, a longer time period would be useful but the text conversation I took my data from was not very long. Also I would ideally use a set amount of alcohol and then analysis conversation at different levels of inebriation, but if the subject was given a set amount and knew it was for study they would presumably alter their speech and so the study would not work.