This post is a high-level overview of the 4 step Data Analytics process that is taught in the WYWM Data Analytics course. In this post I detail what is required at each step of the process, the key points or skills needed, common mistakes to avoid, and which lessons in the Data Analytics Course you should refer back to.
Step 1: Define
What questions do we want to ask our data?
In this stage you need to:
Take the given problem and phrase it in data terms
Understand your dataset
Identify questions for targeted analysis, taking notes of how you might display this data
Note: you will come back to (and sometimes refine) this at all stages of your process.
During the Transform stage you refer back to the questions you’re trying to answer in order to better prep your data
During the Analyse stage you’re looking to generate the insights from the questions of targeted analysis you identified during the Define stage
Key points and skills
Basic quantitative and qualitative analysis
Maintaining an open dialogue with the key stakeholders
Common mistakes to avoid:
Not understanding the required outcome - you should always clarify any points of uncertainty with your key stakeholders
Not spending enough time on this stage - this is arguably the most important step in the process as it is the foundation from which your analysis is built.
Key Lessons:
The data analytics process
Interpreting basic descriptive statistics
Step 2: Transform
Preparing our data for analysis
In this stage, you need to:
Gather your data (from different sources where required)
Merge your data together (where required)
Create and maintain a data dictionary
Record your cleaning steps
Key points and skills:
Merging tables of data together
Power Query - Append
INDEX / MATCH
VLOOKUP
Correcting formatting and normalising data
Dealing with missing or incomplete data
Maintaining a Data Dictionary
updating new columns or fields created (such as calculated fields or using Power Query)
Recording cleaning steps
Common mistakes to avoid:
Segmenting data too early
Incorrectly dealing with missing data - can you use an average of similar values?
Not updating the data dictionary - if you include new fields, you should include them.
Key Lessons:
Cleaning data
Dealing with missing and incorrect values
Text manipulation
Step 3: Analyse
Generating insights or findings
In this stage, you need to:
Generate insights using the questions from your Define stage
Remember that an insight needs to be actionable
Use of exploratory analysis to reveal trends
Spotting patterns
Key points and skills:
Pivot tables and charts
Histograms
Quartiles, percentiles, ranking, classes
Boxplots
Common mistakes to avoid:
Overanalysis or incorrectly targeted analysis - refer to your define stage to keep you on track
Key Lessons:
Exploratory Analysis
Plotting frequency distributions
Calculating and interpreting box-plots and IQR
Step 4: Communicate
Presenting our key findings
In this stage, you need to:
Present your findings in a relevant format, whether that is written, video or audio. Examples include a Powerpoint presentation, a formal meeting, a word document (a template can be found below), an Excel Workbook or Dashboard, or a Power BI report or dashboard.
Ensure your report is tailored for the intended audience. To be valuable, the insights must be understood by the decision maker, who generally isn’t a data analyst.
Keep your presentation and visualisations clean and free of clutter
Key points and skills:
Knowledge of which chart to use to optimally present the insight
Communication
Key Lessons:
Visualisation
Avoiding misleading statistics
Storytelling with Data
Templates
Define, Data Dictionary and Cleaning Steps: Template Link