Preparing for Data Analysis
Preparing for Data Analysis
Recognize data characteristics and choose appropriate analytical methods—from exploratory and descriptive analysis to prediction, inference, performance, gaps, and trends.
Understanding Data Characteristics
Before analysis begins, determine what kind of data you have. The distinction between qualitative and quantitative data affects what can be measured, grouped, transformed, and analyzed.
Quantitative data can be counted, summed, averaged, or measured. Qualitative data describes qualities or categories and can be grouped for analysis and visualization.
Start with the data
Ask how much data exists and whether it is suitable for the intended analysis. Understanding its characteristics helps determine cleaning, transformation, and statistical options.
Then choose the method
The analytical method depends on the available data and the goal of the research. Exploratory analysis is a common starting point before more specific methods.
Quantitative & Qualitative Data
Quantitative data is numerical; qualitative data is categorical. Quantitative data can be further described as discrete or continuous.
| Characteristic | Example | Interpretation |
|---|---|---|
| Discrete quantitative | Number of people in a household | Countable; generally whole-number values. |
| Continuous quantitative | Age, height, weight | Measured; can take increasingly precise values. |
| Qualitative | Favorite color | A category has no inherent numerical magnitude, although category frequencies can be counted. |
Nominal & Ordinal Data
Categorical information can differ in whether a meaningful natural order exists.
Ordinal values have a natural order. Nominal values do not, although an analyst can create a custom sort order when business requirements call for one.
Exploratory & Descriptive Analysis
Exploratory data analysis (EDA) helps reveal the main characteristics of a data set. Descriptive analysis summarizes what the data shows.
Exploratory Data Analysis
Used on data sets to understand their main characteristics, investigate potential relationships, identify correlations, assess possible statistical tests, and determine whether research questions can be addressed.
Descriptive Analysis
Describes observed data using counts, totals, aggregations, and other metrics. It can summarize performance over a short period or across years.
Explore → describe → validate. Learn what is in the data before moving to deeper analysis.
From Description to Action
Different analytical approaches answer different kinds of questions. A useful progression begins with understanding what happened before moving toward prediction, action, or statistical conclusions.
What does the data show?
What value might occur?
What action should the process suggest?
What conclusions are supported?
Basic Analytical Methods
Expand each method to review its purpose and examples from the source material.
Performance, Gap & Trend Analysis
These methods focus respectively on objectives, differences between current and desired states, and movement over time.
Key Takeaways
Countable quantitative values are discrete; measurable quantitative values are continuous. Nominal categories have no natural order; ordinal data does.
EDA identifies the data set's main characteristics. Descriptive methods summarize observed data, predictive methods estimate values, prescriptive methods support action, and inferential methods test relationships and hypotheses.
Performance analysis compares results with objectives, gap analysis compares present and desired states, and trend analysis examines movement over time.
Knowledge Check
Answer all 20 questions. Each question provides immediate feedback.