Purpose of Bias Identification
This page documents the structural biases inherent in Nairobi residential data derived from observable listings. The objective is to clarify how such biases emerge and how they shape what can be observed, without attempting correction, adjustment, or inference.
Bias identification serves as an interpretive boundary. It defines limits on what the dataset can represent rather than proposing remedies or alternative readings.
Publication-Driven Bias
The primary source of bias in listing-based residential data arises from publication behavior. Only residential units that are actively published appear in observable datasets, excluding large portions of stable or informally transacted housing.
This results in overrepresentation of segments where listing is customary and underrepresentation where occupancy is long-term or publication is infrequent.
Form-Related Visibility Bias
Built form introduces a structural bias in observability. Multi-unit developments generate repeated and continuous listing presence due to unit-level turnover, while low-density residential areas appear sporadically or remain absent.
This bias reflects housing typology rather than residential prevalence or significance within the city.
Intermediation and Channel Bias
Residential segments that rely on formal brokerage channels are more likely to appear in listings. Conversely, areas or tenure types that transact through informal networks may be structurally invisible.
Observable data therefore reflects channel usage patterns rather than comprehensive residential coverage.
Labeling and Boundary Bias
District and submarket labels introduce additional bias. Listings may apply inconsistent or overlapping area names, influencing how visibility is grouped and interpreted.
This labeling bias affects apparent distribution without altering underlying residential reality.
Interpretive Limits
These biases do not invalidate the dataset but define its scope. Observable patterns should not be treated as representative samples of Nairobi’s residential system.
Bias identification reinforces the need to separate structural observation from evaluative or directional conclusions.