1 / 10
WGU D204 The Data Analytics Journey Study Guide
1.Analyses for Data Science: Descriptive:: Humans are good at finding
pat- terns, but limited bandwidth – so we need to narrow the data.
Look at the data.
1)Visualize the data – graphs, histograms, bell curve
2)Compute Univariate Descriptive Statistics: mean (average), mode
(most com- mon), median (splits into two equal halves). So ONE
Value.
3)Measures of association: connection between the variables in your
data. Range: high and low, Quartiles, Variance, Standard Deviation,
Correlation coefficients, regression analysis. (this was the click to sales
conversions question)
Must be attentive to outliers, open-ended and undefined scores
(screen data!) Use your words: explain describe
2.Data Analytics Lifecyle per course requirements: Discovery Phase
(Busi- ness understanding)
Data Acquisition
Data Exploration
Predictive
Modeling Data
Mining
2 / 10
Data Reporting & Representation
3.Data Analytics Lifecycle: Discovery: This is where you identify
business needs. Defining the business reason that any analysis is
needed. Working with stakeholders to help them ask better questions
so that both they and you under- stand the outcome. Do you have the
data you need to answer this question? If not, is there another way to
accomplish it with what you do have?
4.Data Analytics Lifecycle: Data Acquisition: Collect/Get Data from
various sources and cleaning it. Getting Data may include creating
SQL queries of data within the tables; or data warehouses. Cleaning is
the most labor-intensive phase (both in time and effort). You may have
to identify outliers here or detect missing values (null values in
columns). An analyst will use SQL,
Python, R, or Excel to perform various data modifications and
transformations. When cleaning is skipped, or ignored, the results from
the analysis may become irrelevant
5.Data Analytics Lifecycle: Data Exploration: Now that the data is
cleaned, you can begin to familiarize yourself with it. You are
beginning to understand the basic nature of data and the
relationships. Making histograms and generally understanding what
is included. Creating bar graphs will give you verification
through visualization. This may include applying a statistical formula to
obtain the avg. temp of a city over the last 50 years. Poor attention to
detail in this phase will give you a lack of insight into the structure of
the data set.
3 / 10
6.Data Analytics Lifecycle: Predictive Modeling: Is where you are taking
those insights and operationalizing them to predict future outcomes
(oil company uses robots to detect corrosion over time – reducing
shutdown/interruptions). You go beyond describing to actual modeling.
Churn Analysis could be performed (eval- uation of a company’s
customer loss repeat in order to reduce it (analyses your product and
how people use it). Common mistake is to develop a model before the
research question is known. Again, Python and R play key roles here.
7.Data Analytics Lifecycle: Data Mining: Find patterns and insights. Find
corre- lations and test hypothesis. Example: data analyst has identified
combinations of sales transactions that frequently occur together in
data over the past 5 years. It is possible in this phase to reduce
significantly the data which results in a sample size that is too small.
Some call this machine learning. Python and R also play key roles
here.
8.Data Analytics Lifecycle: Data Reporting & Representation: Analysts
cre- ates a story to report on data findings. Provide actionable insights
that can inform decision makers; provide conclusion from the analysis
in engaging manner. Ef- fective data reporting is to exclude unrelated
data. Here you may use Tableau or PowerBi.
9.Data Science Pathway (per video):
Planning Wrangling
Modeling
Applying the Model