drt-coursera

[drt-coursera]

Data Analysis Tools - Module4 - by DrT

Testing a Potential Moderator

I examine the relationship between femaleemployrate and urbanrate for 3 different LifeExpectancy categories

Scatterplots:

Summary:

p-value is significant in all of these 3 life exp. categories so LifeExpectancy as a monitor doese have similar effect on femaleemployrate - urbanrate corelation.

[drt-coursera]

Data Analysis Tools - Module3 - by DrT

Pearson Correlation

Calculation:

Scatterplot for the Association Between lifeexpectancy and femaleemployrate:

Scatterplot for the Association Between incomeperperson and femaleemployrate:

Summary:

In the 1st case we have a significant p-value, in the 2nd its not so there is no positive correlation between them.

[drt-coursera]

Data Analysis Tools - Module2 - by DrT

Running a Chi-Square Test of Independence

Alcohol consumption rating:

Set variables and graph:

Contingency table of observed counts in case of 'DRINK_Rating' and 'RankOfUrbanrate':

Contingency table of observed counts in case of 'DRINK_Rating' and 'IncomePerPerson_Rating':

Contingency table of observed counts in case of 'DRINK_Rating' and 'RankOfLifeExpectancy':

Summary:

As a result with groups 'DRINK_Rating' and 'RankOfUrbanrate' we got p=0,04 high and Chi-Square value of 10 high so we can approve the null-hypothesis that there is significant relationship between them.

As a result with groups 'DRINK_Rating' and 'IncomePerPerson_Rating' we got p=0,00038 very low and Chi-Square value of 25.5 high so we can reject the null-hypothesis that there is no significant relationship between them.

As a result with groups 'DRINK_Rating' and 'RankOfLifeExpectancy' we got p=0,00007 very low and Chi-Square value of 24 high so we can reject the null-hypothesis that there is no significant relationship between them.

[drt-coursera]

Data Analysis Tools - Week1 - by DrT

Recoding & creating secondary variables:

Output:

Calculating the F-statistic and associated p value:

Calculating mean & standard deviation:

2nd calculation of the F-statistic and associated p value:

Creating the Multiple Comparison of Means:

Summary:

I've applied the statistical calculations for my database where I analyzed the lifeexpectancy-Urbanrate and lifeexpectancy-femaleemployrate relations. p1=0,0481, p2=0,0108 which means the null hypothesis significantly true stating there's a correlation between them.

[drt-coursera]

Data Management and VisualizationModule 4 - Python My1stProgram

Basic data manipulation for converting ' ' to Nan:

Bar graph for categorical variables:

histogram for quantitative variable:

standard deviation and other descriptive statistics for quantitative variables:

Scatteerplot is created to show data desribed below:

The basic hypothesis is disproved here as it is clearly visible the more the income the most the female employ rate is.

[drt-coursera]

Data Management and Visualization - Module 3

Leave out missing data from the evaluation:

I've created new culomn 'RankOfIncome' for making category for the range of 'incomeperperson':

Ranks:

A: incomeperperson > 10000

B: incomeperperson in 5000 - 10000

C: incomeperperson < 5000

Replace empty cells with category (nan -> 0):

Statistical analysis of variables A, B, C:

Recoding C->A, A->C into a new culomn 'RankOfIncomeReversed':

Statistical analysis of the reversed variables A, C:

I did the same evaluation for 'urbanrate' where I categorized as follows:

urbanrate > 66% : 'A'

urbanrate in 33-66% : 'B'

urbanrate < 33% : 'C'

As 3rd variable I did the same evaluation for 'lifeexpectancy' where I categorized as follows:

lifeexpectancy> 75 : 'A'

lifeexpectancy in 60-75 : 'B'

lifeexpectancy < 60 : 'C'

Summary:

I made evaluation on the income and ranked it to 3 categories. I’ve  categori A a for more than 10,000, B for between 5 and 10,000 and C for less than 5000. I've created similar statistics on urbanrate and lifeexpectancy as described above.

Once I created these variables I could make evaluation on them so I defined the number of the elements which falls into one category and also created the percentage of them.

Next thing I did based on the sample I replaced the categories order (income ) so I replaced A with C so now in the new column A means the less category of the incomes and C means the highest category of the income. Afterwards I’ve calculated the percentage the number of them.

[drt-coursera]

Data Management and Visualization - Module 2

My first program:

Input CSV: Gapminder

My variables:

- urbanrate

- employrate

- internetuserate

My program:

read database:

Print 'urbanrate' quantity, percentage and missing data:

Print 'employrate' quantity, percentage and missing data:

Print 'interentuserate' quantity, percentage and missing data:

Output:

Creating subsets:

Output Subsets:

Copy of subset:

Summary:

Here I created frequency distributions displayed as count & percentage, also displays the missing data.

[drt-coursera]

Data Management and Visualization_Module 1_Getting My Research Project Started

STEP 1: Choose a data set that you would like to work with:

gapminder.csv

STEP 2. Identify a specific topic of interest:

Income per person

STEP 3. Prepare a codebook of your own:

C:\Users\...\gapminder.xlsx

STEP 4. Identify a second topic:

Female employ rate

STEP 5. Add questions/items/variables:

C:\Users\...\gapminder.xlsx

STEP 6. Perform a literature review to see what research has been previously done on this topic:

https://www.jstor.org/stable/2117812

https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1468-232X.1991.tb00775.x

https://www.jstor.org/stable/2808013

STEP 7. Based on your literature review, develop a hypothesis about what you believe the association might be between these topics:

The hypothesis the Income per person is lower the female employ rate is higher .