DTube is a decentralized video streaming app built on top of the STEEM blockchain. It is currently in receipt of delegation from 
But, do we know if all the votes given by DTube are free from bias and independent?
Respiratory:
https://github.com/dtube
Aim of Analysis
The aim of this analysis is to explore if there is a correlation between voting by DTube and other data points contained within the blockchain. Strong correlations would be an indication that DTube voting’s is bias towards that data point, however correlation does not equal causation and a strong positive correlation would be an indicator for further analysis. Whereas no correlation or very weak correlations would suggest independent voting and no further analysis would be required.
What is Correlation?
Correlation is used to test the relationship between two variables, so it measures how things are related. Correlation coefficient is a number between -1 and 1 and measures the strength of the relationship.
This type of analysis is useful when you want to establish if there are possible connections between two variables.
Correlation can be charted as shown below, where one variable is charted against a second. A strong positive correlation will show a line or curve increasing as you move right across the chart as seen in 1 below.
A strong negative correlation will show a line or curve decreasing as you move right across the chart as seen in 4 below.
In addition to charting correlation. One can also calculate the strength of the relationship. This is known as the correlation coefficient.
The coefficient ranges between -1 and +1 and quantifies the direction and strength of the linear association between the two variables. The correlation between two variables can be positive (i.e., higher levels of one variable are associated with higher levels of the other) or negative (i.e., higher levels of one variable are associated with lower levels of the other).
The sign of the correlation coefficient indicates the direction of the association. The magnitude of the correlation coefficient indicates the strength of the association.
For example, a correlation of r = 0.9 suggests a strong, positive association between two variables, whereas a correlation of r = -0.2 suggest a weak, negative association. A correlation close to zero suggests no linear association between two continuous variables.
What is hypothesis testing?
A hypothesis is an educated guess about something in the world around us. It should be testable, either by experiment or observation. Hypothesis testing is the use of statistics to test the results of a survey or experiment to see if you have meaningful results. Basically you are testing whether your results are valid by working out the odds that your results have happened by chance. If your results may have happened by chance, the experiment won’t be repeatable and so has little use.
When carrying out hypothesis testing, a hypothesis question and a null hypothesis are put forward. The null hypothesis is the result of the hypothesis question proving NOT to be true.
Correlation Hypothesis to test for Independence in Voting – Analysis
Test 1:
Null Hypothesis: There is a strong positive correlation between the voting weight used by DTube and the % of app post supported for each creator.
Alternative Hypothesis: There is not a strong positive correlation between the voting weight used by DTube and the % of app post supported for each creator.
Variables:
- average voting weight used for each creator voted on my DTube.
- % of posts made to DTube by each creator that is supported by DTube votes.
If the average weight used increases with the % of posts supported there would be a correlation between the weight used and the overall number of posts DTube support for each creator.
Correlation coefficient = 0.22 – suggests a weak positive correlation. Null hypothesis is therefore rejected.
Test 2:
Null Hypothesis: There is a strong positive correlation between the voting weight used by DTube and the % of post made by each creator on the block that relate to DTube.
Alternative Hypothesis: There is not a strong positive correlation between the voting weight used by DTube and the % of post made by each creator on the block that relate to DTube.
Variables:
- average voting weight used for each creator voted on my DTube.
- % of all posts made to DTube by each creator that via DTube.
If the average weight used increases with the % of posts there would be a correlation.
Correlation coefficient = 0.105 – suggests a weak positive correlation. Null hypothesis is therefore rejected.
Test 3:
Null Hypothesis: There is a strong positive correlation between the voting weight used by DTube and the owned STEEM Power by each creator.
Alternative Hypothesis: There is not a strong positive correlation between voting weight used by DTube and the owned STEEM Power by each creator.
Variables:
- average voting weight used for each creator voted on my DTube.
- SP Owned by each creator.
If the average weight used increases with owned SP there would be a correlation.
Correlation coefficient = 0.1408 – suggests a weak positive correlation. Null hypothesis is therefore rejected.
Test 4:
Null Hypothesis: There is a strong positive correlation between the voting weight used by DTube and the account age for each creator.
Alternative Hypothesis: There is not a strong positive correlation between voting weight used by DTube and the account age for each creator
Data points:
- average voting weight used for each creator voted on my DTube
- account age in days for each creator.
If the average weight used increases with account age in days there would be a correlation
Correlation coefficient = 0.126 – suggests a weak positive correlation. Null hypothesis is therefore rejected.
Test 5:
Null Hypothesis: There is a strong positive correlation between the % of DTube posts supported for each creator and the % of posts made to the blockchain by the creators that are DTube posts.
Alternative Hypothesis: There is not a strong positive correlation between the % of DTube posts supported for each creator and the % of posts made to the blockchain by the creators that are DTube posts.
Variables:
- % of Dtube posts supported for each creator by DTube.
- % of all posts made to the block by each creator that are DTube posts.
If the % of DTube posts supported for each creator by DTube increases with the % of all posts made to the block by each creator that are DTube posts there would be a correlation .
Correlation coefficient = 0.07 – suggests a very weak positive correlation. Null hypothesis is therefore rejected.
Test 6:
Null Hypothesis: There is a strong positive correlation between the % of DTube posts supported for each creator and the SP owned by each creator.
Alternative Hypothesis: There is not a strong positive correlation between the % of DTube posts supported for each creator and the SP Owned by each creator.
Variables:
- % of DTube posts supported for each creator by DTube.
- SP Owned by each creator .
If the % of DTube posts supported for each creator by DTube increases with the SP owned there would be a correlation.
Correlation coefficient = -0.013 – suggests a extremely weak negative correlation. Null hypothesis is therefore rejected.
Test 7:
Null Hypothesis: There is a strong positive correlation between the % of DTube posts supported for each creator and the age of account in days.
Alternative Hypothesis: There is not a strong positive correlation between the % of DTube posts supported for each creator and the age of account in days.
Variables:
- % of DTube posts supported for each creator by DTube.
- The age of account in days.
If the % of DTube posts supported for each creator by DTube increases with the age of account in days there would be a correlation.
Correlation coefficient = -0.019 – suggests an extremely weak negative correlation. Null hypothesis is therefore rejected.
Conclusion
From the 7 hypothesis tests, we have not found any strong correlations between the number of votes or the weight of the vote given and any of the comparable data points. All of the null hypothesis have been rejected. This is a strong indication that DTube vote independently of these variables and are not bias towards account types.
Correlations found are extremely weak and below a threshold that would require further analysis. As all null hypothesis were rejected, no deeper analysis work is required.
The Data and the queries
The data was collected and transformed and modeled using Power BI, however Power BI does not include the functions needed for Statistical analysis like the one above. Therefore, I then transferred the modelled data into Excel to carry out the correlation and produce the visualizations.
All data was collected using SteemSQL held and managed by 
The M queries used to collect and transform the data are:
All Posts
let Source = Sql.Database("vip.steemsql.com", "DBSteem", [Query="select id, author, created, category, total_payout_value from comments#(lf) where CONVERT(DATE,created) BETWEEN '2018-07-01' AND '2018-08-01' and depth = 0"]), #"Changed Type" = Table.TransformColumnTypes(Source,{{"created", type date}})in #"Changed Type"Account Data
let Source = Sql.Database("vip.steemsql.com", "DBSteem", [Query="select name, created, vesting_shares, delegated_vesting_shares, received_vesting_shares from accounts#(lf) "]), #"Replaced Value" = Table.ReplaceValue(Source,"VESTS","",Replacer.ReplaceText,{"vesting_shares", "delegated_vesting_shares", "received_vesting_shares"}), #"Changed Type1" = Table.TransformColumnTypes(#"Replaced Value",{{"vesting_shares", type number}, {"delegated_vesting_shares", type number}, {"received_vesting_shares", type number}}), #"Changed Type" = Table.TransformColumnTypes(#"Changed Type1",{{"created", type date}}), #"Added Custom" = Table.AddColumn(#"Changed Type", "controlling vesting shares", each [vesting_shares]+[received_vesting_shares]-[delegated_vesting_shares]), #"Changed Type2" = Table.TransformColumnTypes(#"Added Custom",{{"controlling vesting shares", type number}}), #"Added Custom1" = Table.AddColumn(#"Changed Type2", "Controlling SP", each [controlling vesting shares]*.000495), #"Added Custom2" = Table.AddColumn(#"Added Custom1", "Owned SP", each [vesting_shares]*.000495), #"Changed Type3" = Table.TransformColumnTypes(#"Added Custom2",{{"Controlling SP", Currency.Type}, {"Owned SP", Currency.Type}})in #"Changed Type3"Posts via APP/TAG
let Source = Sql.Database("vip.steemsql.com", "DBSteem", [Query="select id, author, created, category, total_payout_value, json_metadata from comments#(lf) where CONVERT(DATE,created) BETWEEN '2018-07-01' AND '2018-08-01' and depth = 0"]), #"Changed Type" = Table.TransformColumnTypes(Source,{{"created", type date}}), #"Filtered Rows" = Table.SelectRows(#"Changed Type", each Text.Contains([json_metadata], "Dtube"))in #"Filtered Rows"Votes by APP
let
Source = Sql.Database("vip.steemsql.com", "DBSteem", [Query="select *#(lf)from TXvotes#(lf) where CONVERT(DATE,timestamp) BETWEEN '2018-07-01' AND '2018-08-01'#(lf)and voter = 'dtube' "]), #"Changed Type" = Table.TransformColumnTypes(Source,{{"timestamp", type date}}), #"Added Custom" = Table.AddColumn(#"Changed Type", "% weight", each [weight]/10000), #"Changed Type1" = Table.TransformColumnTypes(#"Added Custom",{{"% weight", Percentage.Type}})in #"Changed Type1"