Topic
Data
What the published distributions say, and how far self-reported figures drift from them.
-
Meta-analysis statistics, start to finish
The statistical machinery behind a pooled size figure: weights, fixed versus random effects, heterogeneity and the intervals that follow.
-
Reading a size study critically
A step-by-step guide to judging any penis-size paper or press story: who was measured, how, by whom, what was pooled and what the number can support.
-
The ratio
Dividing the mean length by the mean girth does not give the mean ratio, and the ratio people quote from the averages is not the ratio most men have.
-
Small samples, wide intervals
A study of a few dozen men can report a mean a centimetre from the pooled estimate purely by sampling; its confidence interval says so if you read it.
-
What was cited before
Before the meta-analysis, single studies served as the reference, each with its own method and sample; knowing them explains older figures still in circulation.
-
What went into the pool
The meta-analysis included only studies where a health professional took the measurement, and that single rule shapes every number it reports.
-
Heaping
Self-reported figures cluster at whole and half units far more than measured ones do, and the pattern itself is evidence of rounding, mostly upward.
-
What normal range means
In a results table, normal range means the middle 95 percent of the sample; it is a description of a distribution, not a judgement, and papers use it that way.
-
Flaccid and erect percentiles
A man's flaccid and erect figures usually land at different percentiles, because the two distributions are shaped differently and flaccid carries more noise.
-
A centimetre near the middle
Because the middle of the distribution is crowded, a centimetre of measurement error near the mean moves a percentile far more than the same centimetre in the tails.
-
Flaccid and erect are loosely related
Flaccid length predicts erect length only weakly at the individual level; the literature reports a correlation, not a rule, and stretched length does better.
-
Forest plot
A forest plot lines up each study's estimate with its interval and the pooled diamond underneath; it is how you see heterogeneity at a glance.
-
Meta-analysis
A meta-analysis pools published studies into one estimate; it inherits their methods and their biases, and it cannot be better than the studies it pools.
-
Z-score to percentile
Subtract the mean, divide by the SD, look up the normal table - three steps take a measurement to a percentile, and each step carries an assumption.
-
Who gets measured
Most measured data come from men attending urology or sexual-health clinics for other reasons, and that sample differs from the population in ways that matter.
-
The online survey problem
Anonymous online surveys combine self-selection, self-measurement and no verification, and the figures they produce sit above every measured study for those reasons.
-
The calculator checklist
A percentile calculator worth using names its reference study, the state and method it assumes, and shows a range for your uncertainty rather than one number.
-
Why the numbers differ between papers
Two clinician-measured studies can report means a centimetre apart because of state, method, population and instrument, before anatomy is involved at all.
-
Bone-pressed in the literature
Not every measured study pressed to the bone, and the pooled figures mix both; a reader placing a bone-pressed reading on them should know that.
-
How to read the histogram
Bin width, sample size and axis choice all change how a histogram looks; here is how to read one without being led by its shape.
-
Fewer girth figures
Fewer studies measured girth, fewer still measured it erect, and the site was not always stated, so the girth pool is thinner than the length pool.
-
Is it a bell curve
The nomograms assume a normal distribution; the measured data are close to it, and where they are not, the tails are the part to distrust.
-
A large sample tightens the mean, not the spread
A big sample pins down the average precisely; it does nothing to narrow the range of individuals, which is set by the standard deviation.
-
Age
The pooled studies are of adult men across a wide age range; what they report about age is limited, and this post says only what the data support.
-
The six-inch number
The round figure that everyone repeats traces back to mid-century self-report and has survived every measured study that came in below it.
-
Which average
Published size figures are means; the median and mode sit close because the distribution is nearly symmetric, but they are not the same number.
-
The girth data
Veale 2015 reports flaccid and erect girth with a tighter spread than length; here are the figures, the sample they come from and what the tighter SD implies.
-
The condom-fit literature
Studies run to size condoms measure erect girth and length self-reported at home, with a fitted product as the reward; they are a separate dataset from the clinical pool.
-
What the top percentile means in centimetres
Under the normal assumption the 99th percentile sits about 2.33 SD above the mean; the Veale figures show how far that is and how few claims survive it.
-
Standard deviation
The SD is the typical distance from the mean, and it is the second number you need before a mean tells you anything about where a figure sits.
-
Citing the paper
The reference is Veale, Miles, Bramley, Muir and Hodsoll, BJU International 2015; here is how to cite it and which figures to attribute to it.
-
Flaccid spread
Across the pooled studies, flaccid length has the largest relative standard deviation of any state, which is the data-side echo of how much it wanders.
-
Shoe size
The shoe-size belief has been tested directly and found no useful relationship; here is what was measured, and why the belief survives anyway.
-
Two calculators, two answers
Online percentile calculators differ because they draw on different studies, different states and sometimes self-report; the disagreement is in the inputs, not the arithmetic.
-
One method change, one shifted mean
Switching a protocol from non-bone-pressed to bone-pressed, or from stretched to erect, moves an entire study's mean by an amount that dwarfs its confidence interval.
-
Ethnicity
Veale 2015 looked for ethnic differences and concluded the pooled data could not support claims either way; that absence of evidence is the honest finding.
-
Three states, three averages
A single table of the Veale means and SDs by state, so the three figures that get confused for each other sit next to each other with their spreads.
-
CI or SD
A confidence interval bounds the mean; a standard deviation describes the individuals. Papers report both and readers routinely swap them.
-
Fewer erect measurements
Far fewer men in the pooled data were measured erect than flaccid or stretched, which is why the erect figures carry wider uncertainty.
-
Reading a girth percentile
Because girth has a smaller standard deviation than length, a few millimetres of girth error move a percentile further than the same error in length would.
-
The studies have error bars too
Clinician measurement is better than self-measurement, not perfect; test-retest figures where reported set a floor under the precision of any pooled number.
-
Which studies get published
Studies with striking findings are likelier to be published and cited; in this literature that favours unusual populations and surprising means.
-
Height
Studies that looked for a link between height and length report at most a weak positive correlation - real in a large sample, useless for predicting any one person.
-
The chart in the article
Percentile charts in magazine features are usually redrawn from a single study or a self-report survey without saying which; a reader should ask.
-
Exclusion criteria
Measured studies exclude men with conditions that affect the measurement, which is correct for their purpose and also trims the tails of the reported distribution.
-
When two clinicians measure
Studies that had two observers measure the same men report good but not perfect agreement, which is the floor of uncertainty in even the best data.
-
Percentile of what
A percentile is only meaningful relative to a named sample; the same figure is a different percentile against a clinic sample, a self-report survey or a single country.
-
The nomogram figures
Veale 2015 published nomograms for flaccid, stretched and erect length and for girth; each is a mean and a spread drawn as a curve, nothing more.
-
The column after the mean
Every measured mean in a paper comes with an SD, SE or CI in the next column, and which one it is changes what the number means.
-
Stretched length across papers
Stretched length depends on how hard the measurer pulls, and no two protocols standardise that identically, so it varies between studies more than erect length does.
-
How rare the extremes are
Under a normal curve, each additional standard deviation from the mean cuts the remaining share sharply; the extremes people talk about are rarer than the talk implies.
-
Two percentiles, one person
Length and girth are only moderately correlated, so most men sit at different percentiles on the two; the same percentile on both is the exception.
-
What the authors say to be careful about
The paper's own limitations section names heterogeneity, clinic-based samples, uneven erect data and unverifiable ethnicity; a reader should carry them along with the means.
-
Nomogram
A nomogram is a chart that lets you read one quantity from another without arithmetic; here it maps a measurement to a percentile.
-
When the reward depends on the number
Some self-measured studies gave participants a product sized to their reported figure; that design changes who takes part and how carefully they measure.
-
Why the papers use centimetres
Clinical studies report in centimetres to one decimal; inch figures in circulation are conversions, and every conversion is a chance for a rounding slip.
-
The average people carry in their heads
Ask people what average is and they name a figure above what clinicians measure; self-report, media selection and viewing angle each push perception the same way.
-
How Kinsey collected size data
Kinsey's figures were self-measured by participants and mailed back on cards, which makes them the founding example of the self-report problem.
-
Designing the honest comparison
The clean test of self-report inflation is the same men measured both ways under one protocol; here is what that design needs and what has come close.
-
Which average did you mean
Flaccid, stretched and erect, bone-pressed or not, clinician-measured or self-reported - each has its own average, and the word alone names none of them.
-
The 2D:4D literature
A handful of studies report a relationship between the second-to-fourth finger ratio and stretched length; it is a research curiosity, not a predictor.
-
One and two standard deviations
Under a normal curve about two-thirds of a sample lie within one SD of the mean and about 95 percent within two; applied to Veale, that gives real ranges.
-
The world map of averages
The country-by-country averages that circulate online mix self-reported surveys with clinical studies of different methods, then rank them to one decimal place.
-
One decimal in the table
A published mean is rounded to match its uncertainty, which is right for the data and wrong to treat as exact; the rounding is a statement about precision.
-
Heterogeneity
When the studies in a pool differ more than chance allows, the pooled mean is an average of different things; Veale 2015 reports this and it should be read.
-
Percentiles
A percentile is the share of the reference sample that measured below a figure; the 50th is the median, the 90th means nine in ten measured less.
-
Who volunteers
Men who agree to be measured for a study are not a random draw, and the direction of that selection is a standing caveat on every reported mean.
-
Why self-reported figures run high
The gap between what people report and what gets measured is consistent, well documented, and mostly not dishonesty. Method explains more of it than motive.
-
The paper that made stretched length a proxy
A small 1996 study measured men flaccid, stretched and erect and found stretched length tracked erect; that finding is why clinics take stretched length today.
-
How a size study is built
Recruitment, exclusion, measurer, instrument, state and reporting: each design choice moves the reported figure, and knowing them is how you read any study.
-
Reading a nomogram without fooling yourself
What a percentile is, how a nomogram is built from a mean and a standard deviation, and how to place a measured figure on it with its uncertainty attached.
-
The Veale meta-analysis, read properly
The paper the whole subject rests on: what it set out to do, what it pooled, what it reports, and where its own authors say to be careful.
-
What the published data actually says
One large meta-analysis, measurements taken by health professionals rather than self-reported. The numbers are lower and the distribution is narrower than almost anyone expects.