CMCE10002 Chap.10 Variation in Cross Section, Time Series and Panel Data
Variation in Cross Section, Time Series and Panel Data
One table, three ways of using it
The week 6 demonstration puts up a single grid of units against periods and reads it three ways. A cross section compares units at one moment, a time series follows one unit across moments, and a panel has both.
Nothing about the file changes; what changes is what one observation has to be.
Naming the shape decides what an average means
A cross-sectional average answers what a typical unit looked like at that moment. A time-series average answers what a typical moment looked like for that unit.
Reporting one and describing the other is invisible in the output, because both are a single number with the same units attached.
Every grouped summary throws one dimension away
Group by the unit and time disappears, leaving differences between units. Group by the period and the units disappear, leaving movement over time.
Grouping reaches values held inside a column and nothing else, so a table with one column per period has to be lengthened before it can be grouped by period at all.
Panels are the shape business data arrives in
Transaction systems record who, what and when, so the natural export carries one row per unit per period whether or not anyone calls it a panel.
The practical consequence is that most analyses begin by deciding which of the two dimensions to give up, and the decision is made silently by whichever grouping is written first. Making it deliberately, and saying which variation was kept, is what separates a defensible summary from a number that simply appeared in the output.
What this chapter covers
- 01
Cross section, time series and panel as three readings of one table
- 02
What one observation is in each of the three shapes
- 03
Grouping by the unit, and the variation that survives
- 04
Grouping by the period, and the variation that survives
- 05
Why a table with one column per period must be lengthened first
Choosing the collapse from the question
- 1Decide the summary for which store performs best.
- 1Decide the summary for whether the business is growing.
- 1Decide what to do when the question needs both dimensions.
Key terms
- Cross section
- Observations on many units at a single point in time, one row per unit.
- Time series
- Observations on one unit across many periods, one row per period.
- Panel data
- Observations on many units across many periods, one row per unit and period.
- Collapsing
- Reducing a table by grouping, which discards the dimension that was grouped over.
Variation in Cross Section, Time Series and Panel Data FAQ
What is the difference between a cross section and a panel?
A cross section holds many units at one moment, so one row is one unit. A panel holds many units across many periods, so one row is one unit in one period. The panel can answer questions about change within a unit that a cross section cannot reach at all.
Can a table be grouped by a period held in the column headings?
No. Grouping reaches values held inside a column and never the headings themselves, so a table with one column per quarter has to be lengthened first. Once the quarter is a value, one row becomes one unit-quarter and the grouping works.
What does an average of a panel actually describe?
It depends on which dimension was grouped away, and the number itself does not say. State the grain of the table it came from and the dimension that was averaged over, because a figure computed across units within one period and a figure computed across periods for one unit look identical.
Which dimension should be grouped away?
The one the question is not comparing. A question about which unit performs best keeps the units and groups time away; a question about whether performance is moving keeps the periods and groups the units away. When both dimensions are being compared, neither collapse is correct and the panel has to stay intact.
Exam move
Write one sentence beside every summary figure naming the grain it came from and the dimension that was averaged away. Neither the code nor the chart supplies that sentence, and short-answer questions frequently ask for exactly it. Rehearse the three questions from the worked example on a table of your own: which unit is best, is the total moving, and is any unit moving against the others.
The third one is the test of whether you have understood why panels exist.
Working through Variation in Cross Section, Time Series and Panel Data in CMCE10002? Sia is AskSia’s AI Statistics tutor — ask any CMCE10002 Variation in Cross Section, Time Series and Panel Data question and get a clear, step-by-step explanation grounded in how CMCE10002 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
