CM1-10 Cross-sectional and longitudinal comparison (cohort and panel comparison; cross-cycle comparison)
Compares different units at one time (cross-sectional) or the same units repeatedly (longitudinal), producing group differences at one time or change paths over time.
What it does
A cross-sectional design compares different units at one time; a longitudinal design observes the same units repeatedly. Artefact: group differences at one time, or change paths; cross-sectional differences may be cohort differences, not change.
When to use it
Use when a user asks how groups differ today or how the same units changed between waves or cycles.
When it does not fit
Cross-sectional differences are read as change, attrition altered a longitudinal sample, or definitions changed between periods.
Result format
The result is either the group differences at one time (cross-sectional) or the change paths of the same units across repeated observations (longitudinal), and it states which design was used. For a cross-sectional result it says that the differences may be cohort differences and not change.
Requirements on the material
| Property | Requirement type | Requirement |
|---|---|---|
| Content type | Always applies | requires=tabular; requires=time series |
| Unit of analysis | Always applies | requires=record; requires=variable |
| Time structure | Applies in some cases | requires=cross-section; requires=repeated cross-section; requires=panel; requires=cohort |
| How the sample was chosen | Applies in some cases | prefers=probability; prefers=census |
| Level of detail available | Always applies | requires=aggregate tables |
| Appraises source credibility | Applies in some cases | no |
Not restricted on: Structure, Variable types, Number of sources, Whether sources can be joined, Amount of material, Known data problems, Producer interest, Source dependence, Origin, Document genre, Language, Personal information, Security classification, Licence.
| Minimum material | At least two groups at one time for a cross-sectional design, or at least two time points for a longitudinal design. |
|---|---|
| Additional sources needed | none |
| Not for (rules) | none |
| Not for (text) | Wrong choice when cross-sectional differences would be read as change, when attrition altered a longitudinal sample, or when definitions changed between periods. |
Choosing and ordering
| Kind of technique | Primary technique, runs in step 5.3 |
|---|---|
| Phase tag | analyse |
| Other phase tags | none |
| Kind of finding | difference; trend |
| Question answered | what; how much |
| Strongest statement it supports | show a link between things |
| Needs as input |
|
| Makes as output |
|
| Must exist first | None |
| Loop target | return to verify and prepare |
Library checks by stakes
Each stakes level includes the checks of the level below it. Checks that need a person apply at high stakes only, so the process leaves them out at lower stakes (decision 38). Repeatability: deterministic.
| Stakes | Check | Person involved | Counts as a different method |
|---|---|---|---|
| exploratory | input validation | none | no |
| decision-informing | input validation | none | no |
| decision-informing | code execution for arithmetic | none | no |
| decision-informing | spot check | optional | no |
| high-stakes | input validation | none | no |
| high-stakes | code execution for arithmetic | none | no |
| high-stakes | spot check | optional | no |
| high-stakes | sensitivity analysis | none | no |
| high-stakes | second technique with different assumptions | none | conditional |
Failure modes and limits
| Model suitability | model plus code |
|---|---|
| Typical model failure | Reads age or group differences at one time as change over time, and ignores attrition and changed definitions. |
| Error sources | input data; data transformation; statistical assumptions |
| Certainty limits | Certainty drops when differences may be cohort differences (indirectness), when attrition changed the sample (risk of bias), or when definitions changed (inconsistency). |
Related techniques
- should follow PE-04 Data profiling and rule-based validation (data audit; data quality assessment; edit checks): definitions and coverage must be checked as consistent across groups and periods
- pairs with HE1-20 Trend and time-series decomposition Also serves: Project forward.: decomposition separates trend from seasonal and irregular parts in repeated series
Evidence and tag basis
Documented. Observational longitudinal studies are argued by some to have less power to detect causation; others disagree. (Wikipedia, Longitudinal study)
Tag basis: input profile inference; selection inference; sequencing inference; robustness inference. All tag values are inference unless a source is named.