# Accuracy, Precision, Truthfulness

When we look at any given mission statistic – e.g. the numbers or percentage of the world that are unreached – we can ask a reflexive question: “Really?!”

Problems with a specific number boil down generally into three categories:

• is the number accurate – that is, is it correct within a specific range, or is it flawed, the methodology for counting off, etc.
• is the number precise – how narrow is the range in which it is correct – to the nearest million, for example, or to the nearest digit
• is the number true – is there any intentional deception, and what is the motivation behind the deception?

Understanding the methodology behind a particular number is important to answering the first two questions.

For example, the definition of “unreached” is a people group lacking a church with the capability of evangelizing the group to its borders without cross-cultural assistance. Unfortunately, there is no numeric measure of this as the term was originally defined. However, two groups that measure this number – Joshua Project and the IMB – have come up with definitions revolving around the percentage of the group that is Christian, and the percentage that is evangelical.

The number in both instances has a fairly simple methodology revolving around one or the other of these numbers being less than a very small percentage (e.g. Christians less than 5% AND evangelicals less than 2% in the case of JP–less than 5% Christian OR less than 2% evangelical within IMB). The assessment of ‘reached’ or ‘unreached’ therefore relies on two specific pieces of data for every people group in the world–and the precision of those two pieces of data is less critical than we might initially think.

This is a yes/no question that is fairly easy to arrive at. The unreached definition only wants to know whether the percentages fall within an admittedly very small range. For most people groups in the world, the answer to that question is fairly clear: they are very obviously either (1) unreached (low % Christian, low % evangelical), (2) unreached by IMB’s definition but not JP (high % Christian, low % evangelical, especially in non-Evangelical majority regions like Europe), or (3) reached (% Christian over 5%, 5 evangelical over 2%).

The unevangelized definition from the World Christian Encyclopedia, on the other hand, incorporates the exact percentage of % Christian into its formula. This makes it more sensitive to the precision of the number. Being over/under the 50% line makes some difference in the summary statistics.

To answer the third question – whether a number is true – requires us to think less about how the data is gathered and more to understand the people and motivations behind the work of gathering and analysis.

In my own experience, I have encountered issues where big national or global numbers in missiology:

• are inaccurate due to lagging (e.g. they haven’t been updated in a long time),
• accused of being too high or low because we (intentionally or otherwise) mistake the methodology (we personally limit “Christian,” for example, to evangelicals, and then hold a research group “wrong” when they don’t use the same definition),
• are too low because they are missing data due to security issues, or even – in a very small handful – intentionally too low to obscure a sensitive security situation on the ground.

But on the big global numbers I have not yet encountered intentional deception.

Now, I know that some ministries and denominations have “evangelistically spoke,” as the euphemism goes, but this is, in the long run, [nearly always?] discovered. It certainly doesn’t endure over the years.

I also know there are instances where large numbers have been or are being debated–are there really that many believers in China?–but the ranges in terms of percentage of population really aren’t that high. For example, in terms of China’s total believer count, I’ve seen arguments about ranges from 2 to 10% or so over the years, but never that China’s believer count is, say, actually 20%. It becomes important in a believer count under 5% would obviously have something to say about unreached. But nowadays–time, again–I don’t know anyone who would claim the number of believers is below the 5% threshold. Time has a way of sorting all these matters out, and researchers are nothing if not patient.

Truthfulness, however, can be a matter of judgment and faith in a source. I know that in some cases, those who say they tell the truth (like myself) have been thought by some to be doing otherwise. In the end, whether someone tells the truth must be left to evidence–or if evidence is unavailable, then left to a judgment of their character. I can only say that for myself, my number one rule is “Non mentior. Never lie. Veritatem quaerite. Seek truth.” Whether you believe me – or anyone – when they say that is a decision the hearer must make.

This entry was posted in Observations. Bookmark the permalink.

### 2 Responses to Accuracy, Precision, Truthfulness

1. I have been concerned about religious statistics for a long time, especially those of evangelical Christians. To me the issue is about what is a Christian and whether that person is living out their faith as well as assuming everyone that who goes to an evangelical church is an evangelical and/or the whole families are Christians. I also wonder if churches ‘inflate’ figures. On the other hand, Christians who cannot get to church or even attend if they wanted to (because of persecution or disability or other reasons) are usually not counted. We need to find a better way to accurately count evangelicals ad not simply use so called ‘statistical’ ‘rules’ or ‘laws’.

• Justin Long says:

Doing any sort of counting is very difficult. Global counts aren’t done on an annual basis due to the complexity and time and cost. In many places, there is not consistent agreement on what counts as an evangelical. Global counts are done on the basis of how individual denominations define and make counts, and every denomination has its own issues with their counts. And obviously precise counts vary from week to week, month to month, year to year. All counts for these and other reasons have to be rounded estimates that show orders of magnitude and trend lines – religious demography is often more like a fuzzy Monet than a black-and-white photograph!