Proper interpretation of non-differential misclassification effects: expectations vs observations
The DOI matches its Crossref record; 26 of the 26 references it cites were found in a registry or an index.
Abstract
BACKGROUND: Many investigators write as if non-differential exposure misclassification inevitably leads to a reduction in the strength of an estimated exposure-disease association. Unfortunately, non-differentiality alone is insufficient to guarantee bias towards the null. Furthermore, because bias refers to the average estimate across study repetitions rather than the result of a single study, bias towards the null is insufficient to guarantee that an observed estimate will be an underestimate. Thus, as noted before, exposure misclassification can spuriously increase the observed strength of an association even when the misclassification process is non-differential and the bias it produced is towards the null. METHODS: We present additional results on this topic, including a simulation study of how often an observed relative risk is an overestimate of the true relative risk when the bias is towards the null. RESULTS: The frequency of overestimation depends on many factors: the value of the true relative risk, exposure prevalence, baseline (unexposed) risk, misclassification rates, and other factors that influence bias and random error. CONCLUSIONS: Non-differentiality of exposure misclassification does not justify claims that the observed estimate must be an underestimate; further conditions must hold to get bias towards the null, and even when they do hold the observed estimate may by chance be an overestimate.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Measurement Error in Nonlinear Models
Carroll, R. J., Ruppert, D., & Stefanski, L. A. (1995) · Springer eBooks · doi:10.1007/978-1-4899-4477-1 · Cited by 1,749
Risk Analysis: A Quantitative Guide
Vose, D. (2000) · Cited by 1,747
Uncertainty: A Guide to Dealing with Uncertainty in Quantitative Risk and Policy Analysis
Morgan, M. G., & Henrion, M. (1990) · Cambridge University Press eBooks · doi:10.1017/cbo9780511840609 · Cited by 1,711
Balfour, J. L. (2002) · American Journal of Epidemiology · doi:10.1093/aje/155.6.507 · Cited by 595
Multiple-bias modelling for analysis of observational data
Greenland, S. (2005) · Journal of the Royal Statistical Society Series A (Statistics in Society) · doi:10.1111/j.1467-985x.2004.00349.x · Cited by 497
DOES NONDIFFERENTIAL MISCLASSIFICATION OF EXPOSURE ALWAYS BIAS A TRUE EFFECT TOWARD THE NULL VALUE?
DOSEMECI, M., WACHOLDER, S., & LUBIN, J. H. (1990) · American Journal of Epidemiology · doi:10.1093/oxfordjournals.aje.a115716 · Cited by 404
Misclassification in 2 X 2 Tables
Bross, I. (1954) · Biometrics · doi:10.2307/3001619 · Cited by 379
Differential Misclassification Arising from Nondifferential Errors in Exposure Measurement
Flegal, K. M., Keyl, P. M., & Nieto, F. J. (1991) · American Journal of Epidemiology · doi:10.1093/oxfordjournals.aje.a116026 · Cited by 281
Semi-Automated Sensitivity Analysis to Assess Systematic Errors in Observational Data
Lash, T. L., & Fink, A. K. (2003) · Epidemiology · doi:10.1097/01.ede.0000071419.41011.cf · Cited by 250
Steenland, K. (2004) · American Journal of Epidemiology · doi:10.1093/aje/kwh211 · Cited by 208
Cited by
Posner, K., Oquendo, M. A., Gould, M., Stanley, B., & Davies, M. (2007) · American Journal of Psychiatry · doi:10.1176/ajp.2007.164.7.1035 · Cited by 1,025
The Impact of Residual and Unmeasured Confounding in Epidemiologic Studies: A Simulation Study
Fewell, Z., Davey Smith, G., & Sterne, J. A. C. (2007) · American Journal of Epidemiology · doi:10.1093/aje/kwm165 · Cited by 734
“a bias towards the null (4), the effects of measurement error”
BackgroundMeasurement error and the replication crisis
Loken, E., & Gelman, A. (2017) · Science · doi:10.1126/science.aal3618 · Cited by 628
Meeting Report: Atmospheric Pollution and Human Reproduction
Slama, R., Darrow, L., Parker, J., Woodruff, T. J., Strickland, M., Nieuwenhuijsen, M., Glinianaia, S., Hoggatt, K. J., Kannan, S., Hurley, F., et al. (2008) · Environmental Health Perspectives · doi:10.1289/ehp.11074 · Cited by 322
Arfè, A., Scotti, L., Varas-Lorenzo, C., Nicotra, F., Zambon, A., Kollhorst, B., Schink, T., Garbe, E., Herings, R., Straatman, H., et al. (2016) · BMJ · doi:10.1136/bmj.i4857 · Cited by 290
A method to automate probabilistic sensitivity analyses of misclassified binary variables
Fox, M. P., Lash, T. L., & Greenland, S. (2005) · International Journal of Epidemiology · doi:10.1093/ije/dyi184 · Cited by 288
Self-citationPurpose in life and reduced incidence of stroke in older adults: 'The Health and Retirement Study'
Kim, E. S., Sun, J. K., Park, N., & Peterson, C. (2013) · Journal of Psychosomatic Research · doi:10.1016/j.jpsychores.2013.01.013 · Cited by 274
Blair, A., Stewart, P., Lubin, J. H., & Forastiere, F. (2007) · American Journal of Industrial Medicine · doi:10.1002/ajim.20281 · Cited by 251
Morgan, R. L., Thayer, K. A., Santesso, N., Holloway, A. C., Blain, R., Eftim, S. E., Goldstone, A. E., Ross, P., Ansari, M., Akl, E. A., et al. (2019) · Environment International · doi:10.1016/j.envint.2018.11.004 · Cited by 248
“If non-differential, the exposure misclassification will usually bias associations to the null, although the final impact on the observed relative risk is also dependent on other factors (Jurek et al., 2005).”
BackgroundLivesey, G., Taylor, R., Livesey, H. F., Buyken, A. E., Jenkins, D. J. A., Augustin, L. S. A., Sievenpiper, J. L., Barclay, A. W., Liu, S., Wolever, T. M. S., et al. (2019) · Nutrients · doi:10.3390/nu11061280 · Cited by 245
“Under conditions of low or nil random error (Table 8), it is more likely that the risk relations are underestimated than overestimated due to such misclassification [86].”
BackgroundTradeoffs between accuracy measures for electronic health care data algorithms
Chubak, J., Pocobelli, G., & Weiss, N. S. (2012) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2011.09.002 · Cited by 215
“Because bias is an average, chance alone may cause results from an individual study to be in the opposite direction of the expected bias [45,55,56].”
ResultChang, E. T., Adami, H.-O., Boffetta, P., Cole, P., Starr, T. B., & Mandel, J. S. (2014) · Critical Reviews in Toxicology · doi:10.3109/10408444.2014.905767 · Cited by 180
“Additional conditions must be satisfi ed for the bias to be toward the null, and even when all such conditions are met, a given estimate may by chance be biased away from the null (Jurek et al. 2005, Jurek et al. 2008).”
BackgroundReflection on modern methods: five myths about measurement error in epidemiological research
van Smeden, M., Lash, T. L., & Groenwold, R. H. H. (2020) · International Journal of Epidemiology · doi:10.1093/ije/dyz251 · Cited by 173
“In a particular dataset with measurement error, exposure effects may be overestimated only due to sampling variability, illustrations of which are found in Hutcheon et al. and Jurek et al.22,43 Hence, a defining characteristic of the iron law is that it applies to averages of exposure effect estimates, e.g. after many…”
BackgroundRandom measurement error: Why worry? An example of cardiovascular risk factors
Brakenhoff, T. B., van Smeden, M., Visseren, F. L. J., & Groenwold, R. H. H. (2018) · PLOS ONE · doi:10.1371/journal.pone.0192298 · Cited by 52
Implications of the Use of Algorithmic Diagnoses or Medicare Claims to Ascertain Dementia
Power, M. C., Gianattasio, K. Z., & Ciarleglio, A. (2020) · Neuroepidemiology · doi:10.1159/000510753 · Cited by 21
Pasqual, E., Turner, M. C., Gracia-Lavedan, E., Casabonne, D., Benavente, Y., Chef, I. T., Maynadié, M., Cocco, P., Staines, A., Foretova, L., et al. (2020) · PLOS ONE · doi:10.1371/journal.pone.0235658 · Cited by 8
“Even when non-differential between cases and controls, recall error can affect risk estimates, though the direction of the effect is unpredictable [47].”
Background
Metadata
- Authors
- Jurek, A. M., Greenland, S., Maldonado, G., & Church, T. R.
- Venue
- International Journal of Epidemiology
- Publisher
- Oxford University Press (OUP)
- Published
- 2005-03-31
- Volume
- 34
- Issue
- 3
- Pages
- 680-687
- ISSN
- 0300-5771 (ISSN-L)
- 1464-3685
- OpenAlex
- W2143637946
- Cited by
- 410
- References
- 26
- FWCI
- 10.92
- Type
- journal article
- Language
- en
- Open access
- closed
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.