Advanced search
Add to list

Assessing small sample bias in coordinate based meta-analyses for fMRI

Freya Acar (UGent) , Ruth Seurinck (UGent) and Beatrijs Moerkerke (UGent)
(2016)
Author
Organization
Abstract
Meta-analyses for fMRI are booming but remain challenging, given the complex structure of the data and the often censored reporting of results. Because many fMRI studies only report locations of peak voxels that survive a statistical threshold, coordinate-based methods (CBMA), such as ALE [1,2,3], have been developed. Meta-analyses are in general prone to publication bias or the systematic difference between results of published and unpublished studies, but fMRI studies may be particularly susceptible to small sample bias. fMRI studies with small sample sizes may tend to employ more lenient thresholding to compensate for underpowered tests. In classical meta-analyses this is assessed by regressing the observed effect sizes of studies on the sample sizes [4]. The limited amount of information that serves as input for CBMA prohibits the use of these classical techniques to assess and correct for publication bias. In this study, we explore patterns in ALE results that are observed under small sample bias scenarios. The presence of small sample bias in the form of lenient thresholding alters the ALE results and reveals distinct patterns in the regression of sample sizes on study contribution, with smaller slopes in meta-analyses that suffer from small sample bias. [1] Eickhoff, S.B., Laird, A.R., Grefkes, C., Wang, L.E., Zilles, K., and Fox, P.T. (2009), ‘Coordinate-based activation likelihood estimation meta-analysis of neuroimaging data: A random-effects approach based on empirical estimates of spatial uncertainty’, Human Brain Mapping, vol. 30, pp. 2907-2926. [2] Eickhoff, S.B., Bzdok, D., Laird, A.R., Kurth, F., and Fox, P.T. (2012), ‘Activation likelihood estimation revisited’, NeuroImage, vol. 59, pp. 2349-2361. [3] Turkeltaub, P.E., Eickhoff, S.B., Laird, A.R., Fox, M., Wiener, M., and Fox, P. (2012), ‘Minimizing within-experiment and within-group effects in activation likelihood estimation meta-analyses’, Human Brain Mapping, vol. 33, pp. 1-13. [4] Egger, M., Davey Smith, G., Schneider, M., and Minder, C. (1997), ‘Bias in meta-analysis detected by a simple, graphical test’, British Medical Journal, vol. 315, pp. 629-634.
Keywords
fMRI, meta-analysis, publication bias

Citation

Please use this url to cite or link to this publication:

MLA
Acar, Freya, Ruth Seurinck, and Beatrijs Moerkerke. “Assessing Small Sample Bias in Coordinate Based Meta-analyses for fMRI.” 2016. Print.
APA
Acar, F., Seurinck, R., & Moerkerke, B. (2016). Assessing small sample bias in coordinate based meta-analyses for fMRI. Presented at the Channel Network Conference of the International Biometric Society.
Chicago author-date
Acar, Freya, Ruth Seurinck, and Beatrijs Moerkerke. 2016. “Assessing Small Sample Bias in Coordinate Based Meta-analyses for fMRI.” In .
Chicago author-date (all authors)
Acar, Freya, Ruth Seurinck, and Beatrijs Moerkerke. 2016. “Assessing Small Sample Bias in Coordinate Based Meta-analyses for fMRI.” In .
Vancouver
1.
Acar F, Seurinck R, Moerkerke B. Assessing small sample bias in coordinate based meta-analyses for fMRI. 2016.
IEEE
[1]
F. Acar, R. Seurinck, and B. Moerkerke, “Assessing small sample bias in coordinate based meta-analyses for fMRI,” presented at the Channel Network Conference of the International Biometric Society, Hasselt, 2016.
@inproceedings{8520095,
  abstract     = {Meta-analyses for fMRI are booming but remain challenging, given the complex structure of the data and the often censored reporting of results. Because many fMRI studies only report locations of peak voxels that survive a statistical threshold, coordinate-based methods (CBMA), such as ALE [1,2,3], have been developed. Meta-analyses are in general prone to publication bias or the systematic difference between results of published and unpublished studies, but fMRI studies may be particularly susceptible to small sample bias. fMRI studies with small sample sizes may tend to employ more lenient thresholding to compensate for underpowered tests. In classical meta-analyses this is assessed by regressing the observed effect sizes of studies on the sample sizes [4]. The limited amount of information that serves as input for CBMA prohibits the use of these classical techniques to assess and correct for publication bias. In this study, we explore patterns in ALE results that are observed under small sample bias scenarios. The presence of small sample bias in the form of lenient thresholding alters the ALE results and reveals distinct patterns in the regression of sample sizes on study contribution, with smaller slopes in meta-analyses that suffer from small sample bias. 

[1] Eickhoff, S.B., Laird, A.R., Grefkes, C., Wang, L.E., Zilles, K., and Fox, P.T. (2009), ‘Coordinate-based activation likelihood estimation meta-analysis of neuroimaging data: A random-effects approach based on empirical estimates of spatial uncertainty’, Human Brain Mapping, vol. 30, pp. 2907-2926.
[2] Eickhoff, S.B., Bzdok, D., Laird, A.R., Kurth, F., and Fox, P.T. (2012), ‘Activation likelihood estimation revisited’, NeuroImage, vol. 59, pp. 2349-2361.
[3] Turkeltaub, P.E., Eickhoff, S.B., Laird, A.R., Fox, M., Wiener, M., and Fox, P. (2012), ‘Minimizing within-experiment and within-group effects in activation likelihood estimation meta-analyses’, Human Brain Mapping, vol. 33, pp. 1-13.
[4] Egger, M., Davey Smith, G., Schneider, M., and Minder, C. (1997), ‘Bias in meta-analysis detected by a simple, graphical test’, British Medical Journal, vol. 315, pp. 629-634.},
  author       = {Acar, Freya and Seurinck, Ruth and Moerkerke, Beatrijs},
  keywords     = {fMRI,meta-analysis,publication bias},
  location     = {Hasselt},
  title        = {Assessing small sample bias in coordinate based meta-analyses for fMRI},
  year         = {2016},
}