Purpose of Bias Disclosure
This page centralizes disclosure of known biases and structural limitations inherent in listing-based residential datasets used for Dar es Salaam. Its purpose is to define interpretive boundaries clearly and to prevent unsupported conclusions derived from partial visibility.
Visibility Bias
The dataset reflects only residential properties that are actively published and visible on selected platforms at a specific point in time. Properties that are unlisted, informally exchanged, or circulated through non-digital channels are structurally excluded, regardless of their prevalence or significance.
Contributor and Platform Dependence
Listing content depends on contributor behavior, platform rules, and interface design. Attribute completeness, categorization, and geographic labeling vary based on how contributors choose to disclose information and how platforms constrain available options.
Geographic and Administrative Bias
Location references in listings rely on administrative boundaries, neighborhood names, or generalized descriptors selected by contributors. This can lead to uneven spatial representation that reflects documentation practices rather than actual residential distribution.
Temporal Snapshot Limitation
The dataset represents a single visibility snapshot rather than a continuous or historical record. Listing appearance or disappearance reflects publication cycles and moderation practices, not changes in residential structure or housing conditions.
Non-Representativeness of Housing Stock
Because the dataset captures visibility rather than existence, it cannot be treated as representative of residential stock, occupancy, or composition. Absence of data should not be interpreted as absence of residential assets.