View article

An evolving literature evaluates the inferential and behavioral implications of measurement error (ME) in agricultural data. We synthesize findings on the nature and sources of ME and potential remedies. We provide practical guidance for choosing among alternative approaches for detecting, obviating, or correcting for alternative sources of ME, as these have different behavioral and inferential implications. Some ME biases statistical inference and thus may require econometric correction. Other types of ME may affect (and shed light on) farmers’ decision-making processes even if farmers’ responses are objectively incorrect. Where feasible, collecting both self-reported and objectively measured data for the same variable may enrich understanding of policy-relevant agricultural and behavioral phenomena.