Statistical analysis needed, data science

Please read the description carefully. You will find the needed databases attached. Look at them, and when contacting me with your proposal please describe briefly what methodology you would use. Please avoid generic descriptions / I would not read at your CV, I really want someone, who would do the job. The data sets are extremelly smal. I am not looking to develop a trading model, this is needed for a thesis during my bachelor study. The aim of the analysis is to find the interdependencies between the closing price of a tradable asset and the sentiment of news. The first document has the following variables: Closing price - the closing the price of the asset, on weekly basis, in USD. For example - 1000USD Change - change in % from previous week, f.e. +2.0% or -0.65%. The period is about 155 weeks - mid january 2015 till last week of 2017. Second file includes 244 manually annotated articles from certain news portal. Sometimes a week or two weeks in a row there are no news. If that is too problematic for the calculation / if it gets too complicated, I could extract monthly information. I can also get average price for the set period of time, instead of the closing price. You let me know what makes more sense / what would be easier for you to do the job. The second file has the following variables (1)TITLE - Title of the article (2)DATE - The date of publishing in the format dd/mm/yyyy (3)TITLE SENTIMENT - Sentiment of the titles - POSITIVE / NEGATIVE / NEUTRAL (4)ARTICLE SENTIMENT - Sentiment of the article itself - POSITIVE / NEGATIVE / NEUTRAL (5)BUBBLE IMPLICATION - If the article suggests, that the asset is in a bubble - YES / NO (6)LENGTH WORDS - Length of the article in absolute number of words (7)LENGTH CHARACTERS - Length of the article in characters without blanks / with blanks (not relevant for the job) (8)NAG - The type of the article - NEWS, STORY, ANALYSIS, INTERVIEW, OTHER (9)ALPHAVILLE - YES / NO - If the article is part of [login to view URL] specialized financial analyses (10)PREMIUM - Y - If the article is part of their premium content, again not relevant for the job. There are a few hypotheses (5-6). As stated earlier, the aim is not to achieve a trading model (the bachelor thesis is not in finance, but in social science), but rather to be able to conclude the following (just as an example, real conclusion depends on the results). - Sentiment of the news(4) shifts to negative with the price going up. Here the aim is to conclude if the sentiment moves the price, or if the price moves the sentiment, i.e. what moves first (time-analysis?) -There is tendency for the negative(4) analyses(8 / NAG, the type of the news) to get longer (6 - length of the article in words) with the price going up. - News (All news, not depending on type) hinting that the asset might be in a bubble (5) become more frequent with the price going up. - With the price going up the news (no matter of their type) become more frequent / more negative / longer (1 variable from these.) As a whole, I am interested in such type of interdependencies. I am not too much into statistics, so I will need 1-2 sentences only explaining the results. From there on, I will work on extending the descriptions. I am down to discuss some changes or propositions, which would not harm the idea, but will make your job easier / faster. Please contact me if interested with your price quotation and when it could be done. Please do not contact me if the price exceeds my tight budget by too much. Best regards.