Testing for the presence of measurement error in Stata

Lee, Young Jun; Wilhelm, Daniel

doi:10.1177/1536867x20931002

Cited by 6 publications

(4 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Regardless of the causes of the existence and persistence of misperceptions, failure to account for misperceptions could lead to biased and inconsistent price premium estimates. To this end, we examine whether the remaining quality misperception is severe enough to distort the price premia estimates reported earlier in Table 2 using non-parametric hypothesis testing approaches ( Lee and Wilhelm, 2020 ; Wossen et al, 2022 ). Non-parametric test results, based on the Cramer-von Mises test statistics ( CvM ) and Kolmogorov-Smirnov test statistics ( KS ), consistently reject the null of no misperception post-reveal at all reasonable significance levels (see Appendix Table A6 ).…”

Section: Resultsmentioning

confidence: 99%

Estimating seed demand in the presence of market frictions: Evidence from an auction experiment in Nigeria

Wossen,

Spielman,

Alene

et al. 2024

Journal of Development Economics

View full text Add to dashboard Cite

Section: Resultsmentioning

confidence: 99%

Estimating seed demand in the presence of market frictions: Evidence from an auction experiment in Nigeria

Wossen,

Spielman,

Alene

et al. 2024

Journal of Development Economics

View full text Add to dashboard Cite

“…We find that the coefficient of Z is statistically significant at the 5% level (t-statistic of 2.21). Therefore, with the assumption of linearity, we reject the null of no ME in self-reported education (Hausman (1978) and Lee and Wilhelm (2020)). This finding is consistent with the possibility that the nonparametric test has low power because of the small sample size and thus fails to reject, but the test imposing linearity of the conditional expectation is more powerful and thus does reject.…”

Section: Us Twins Datamentioning

confidence: 88%

Powerful t-tests in the presence of nonclassical measurement error

2021

View full text Add to dashboard Cite

This paper proposes a powerful alternative to the t-test in linear regressions when a regressor is mismeasured. We assume there is a second contaminated measurement of the regressor of interest. We allow the two measurement errors to be nonclassical in the sense that they may both be correlated with the true regressor, they may be correlated with each other, and we do not require any location normalizations on the measurement errors. We propose a new maximal t-statistic that is formed from the regression of the outcome onto a maximally weighted linear combination of the two measurements. Critical values of the test are easily computed via a multiplier bootstrap. In simulations, we show that this new test can be significantly more powerful than t-statistics based on OLS or IV estimates. Finally, we apply the proposed test to the studies of returns to education based on twins data from the US and the UK. With our maximal t-test, we are able to discover statistically significant returns to education when standard t-tests do not.

show abstract

“…Therefore, with the assumption of linearity, we still suspect the presence of measurement errors while the tests do not strongly reject the null of no ME. (Hausman (1978) and Lee and Wilhelm (2020)). This finding is consistent with the possibility that the nonparametric test has low power because of the small sample size and thus fails to reject, but the test imposing linearity of the conditional expectation is more powerful.…”

Section: Empirical Illustration: Returns To Schoolingmentioning

confidence: 98%

Powerful t-tests in the presence of nonclassical measurement error

2023

View full text Add to dashboard Cite

This paper proposes a powerful alternative to the t-test of the null hypothesis that a coefficient in linear regression is equal to zero when a regressor is mismeasured. We assume there are two contaminated measurements of the regressor of interest. We allow the two measurement errors to be nonclassical in the sense that they may both be correlated with the true regressor, they may be correlated with each other, and we do not require any location normalizations on the measurement errors. We propose a new maximal t-statistic that is formed from the regression of the outcome onto a maximally weighted linear combination of the two measurements. Critical values of the test are easily computed via a multiplier bootstrap. In simulations, we show that this new test can be significantly more powerful than t-statistics based on OLS or IV estimates. Finally, we apply the proposed test to a study of returns to education based on twins data from the UK. With our maximal t-test, we can discover statistically significant returns to education when standard t-tests do not.

show abstract

Testing for the presence of measurement error in Stata

Cited by 6 publications

References 53 publications

Estimating seed demand in the presence of market frictions: Evidence from an auction experiment in Nigeria

Estimating seed demand in the presence of market frictions: Evidence from an auction experiment in Nigeria

Powerful t-tests in the presence of nonclassical measurement error

Powerful t-tests in the presence of nonclassical measurement error

Contact Info

Product

Resources

About