Visual bug in displaying LaTeX equations.

[hufengling]

There appears to be a visual bug when displaying certain lines of LaTex, seen in screenshot below, found in linked paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5393436/. Visual bug is reproducible for same equations (but not every equation) in macOS Chrome and Opera, but not present in Safari. Bug is not seen in PDF or classic view.

[kolotev]

May I ask for your platform/browser/version? I tried Firefox & Safari on Mac and found no problem so far. By the way, the source of that math is not LaTeX, but MathML. It is rendered with help of MathJax.

Thank you for reporting.

[hufengling]

Noticed on Google Chrome (Version 80.0.3987.132) and Opera (Version:67.0.3575.53). Currently using macOS Sierra (Version 10.12.6). Thanks for the quick response and MathML clarification!

[kolotev]

Seems like a MathJax issue, when I change the geometry of the browser sometimes it renders better sometimes worse. Here is the markup of the HTML page related to these paragraph

<p id="__p13" class="p p-last">In the second stage of MI, the intended analysis is performed on each of the <em>m</em> complete datasets, in this case model (1). Let <span id="ilm9-0962280216683570"><math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="mml-math9-0962280216683570" overflow="scroll"><mrow><msub><mrow><mover><mrow><mi>&#x003b8;</mi></mrow><mi>^</mi></mover></mrow><mrow><mi>j</mi></mrow></msub></mrow></math></span> denote the estimate of <span id="ilm10-0962280216683570"><em>&beta;</em><sub>1</sub></span> from the <em>j</em>th imputed dataset and <em>W<sub>j</sub></em> the corresponding variance estimate. Using Rubin's rules,<sup><a href="#bibr6-0962280216683570" rid="bibr6-0962280216683570" class=" bibr popnode">6</a></sup> the combined MI treatment effect estimate <span id="ilm11-0962280216683570"><math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="mml-math11-0962280216683570" overflow="scroll"><mrow><mover><mrow><mi>&#x003b8;</mi></mrow><mi>^</mi></mover></mrow></math></span> is calculated as the mean of the <em>m</em> estimates, i.e. <span id="ilm12-0962280216683570"><math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="mml-math12-0962280216683570" overflow="scroll"><mrow><mover><mrow><mi>&#x003b8;</mi></mrow><mi>^</mi></mover><mo>=</mo><mn>1</mn><mo>/</mo><mi>m</mi><munderover><mo>&#x02211;</mo><mrow><mi>j</mi><mo>=</mo><mn>1</mn></mrow><mrow><mi>m</mi></mrow></munderover><msub><mrow><mover><mrow><mi>&#x003b8;</mi></mrow><mi>^</mi></mover></mrow><mrow><mi>j</mi></mrow></msub><mo>.</mo></mrow></math></span> The variance is given by <span id="ilm13-0962280216683570"><math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="mml-math13-0962280216683570" overflow="scroll"><mrow><mtext>var</mtext><mrow><mo>(</mo></mrow><mover><mrow><mi>&#x003b8;</mi></mrow><mi>^</mi></mover><mrow><mo>)</mo></mrow><mo>=</mo><mi>W</mi><mo>+</mo><mi>B</mi><mrow><mo>(</mo></mrow><mn>1</mn><mo>+</mo><mn>1</mn><mo>/</mo><mi>m</mi><mrow><mo>)</mo></mrow></mrow></math></span>, where <span id="ilm14-0962280216683570"><math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="mml-math14-0962280216683570" overflow="scroll"><mrow><mi>W</mi><mo>=</mo><mn>1</mn><mo>/</mo><mi>m</mi><munderover><mo>&#x02211;</mo><mrow><mi>j</mi><mo>=</mo><mn>1</mn></mrow><mrow><mi>m</mi></mrow></munderover><msub><mrow><mi>W</mi></mrow><mrow><mi>j</mi></mrow></msub></mrow></math></span> is the average within-imputation variance and <span id="ilm15-0962280216683570"><math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="mml-math15-0962280216683570" overflow="scroll"><mrow><mi>B</mi><mo>=</mo><mo stretchy="false">(</mo><mi>m</mi><mo>-</mo><mn>1</mn><msup><mo stretchy="false">)</mo><mrow><mo>-</mo><mn>1</mn></mrow></msup><munderover><mo>&#x02211;</mo><mrow><mi>j</mi><mo>=</mo><mn>1</mn></mrow><mrow><mi>m</mi></mrow></munderover><mo stretchy="false">(</mo><msub><mrow><mover><mrow><mi>&#x003b8;</mi></mrow><mi>^</mi></mover></mrow><mrow><mi>j</mi></mrow></msub><mo>-</mo><mover><mrow><mi>&#x003b8;</mi></mrow><mi>^</mi></mover><msup><mo stretchy="false">)</mo><mrow><mn>2</mn></mrow></msup></mrow></math></span> the between imputation variance. Hypothesis tests and confidence intervals can be obtained using a <em>t</em>-distribution with <span id="ilm16-0962280216683570"><em>v</em> =&nbsp;(<em>m</em> -&nbsp;1)[1 +&nbsp;<em>W</em>/(1 +&nbsp;<em>m</em><sup>-1</sup>)<em>B</em>]<sup>2</sup></span> degrees of freedom.</p></div><div id="sec6-0962280216683570" class="sec sec-last"><h3 id="sec6-0962280216683570title">

It is probably a time to try a newer version of MathJax with hope, that it was improved.
Unfortunately, I can not provide you with estimate on time when that would happen, busy with other tasks.