Conversion therapy, religious belief and the Church – Part 3

Ozanne Foundation

I’ve also got a guest post about this topic on the Archbishop Cranmer blog here.

HARK the herald activists sing

Back in December 2018 I wrote two posts (Part 1 and Part 2) about the Ozanne Foundation’s Faith and Sexuality Survey, a confused piece of research that had already set itself up for problems because the questions in the survey could not adequately address the aims it proclaimed.

So how have the ‘findings’ turned out at General Synod this month?  Well, as I predicted, a number of cause-and-effect claims that cannot be supported by the data have already been made by prominent figures (Bishop Bayes of Liverpool and Jayne Ozanne).

Research is fraught with pitfalls.  When reading about a piece of research, especially if you don’t know how to actually do research effectively, it helps to consider a few the old jokes that circulate in professional research circles:

‘You are completely free to carry out your study in whatever way you want, as long as you come to these conclusions.’

‘While you’re doing research, bear in mind that there are only two kinds of facts relevant to this topic – those that support our position and those that are inconclusive.’

‘I can’t help noticing that your recent research reports are getting more and more autobiographical, Fred’.

‘Did you read my paper about confirmation bias?’  ‘Yes but it only endorsed what I already thought’.

‘Evidence shows that people who live longest have more birthdays than everyone else’.

Are you getting the idea?  Well read on and inwardly digest.

Confirmation bias and flawed reasoning

What we see in the results and public announcements about the Faith and Sexuality Survey is a classic case of confirmation bias.

That’s a tendency we all have to look for – and find – evidence that supports our own pre-existing values or beliefs.  It’s a failure of reasoning.  I could be said to be biased because I used to direct a Master’s programme in research methods which was used as a test-bed for postgraduate students wishing to do a PhD.  I am now ‘programmed’ to look for faults in any research I read.  I try to minimise my bias by also looking for the positive aspects of the research that I read and by being aware that I could be making unreasonable criticisms. I also ask people to reach their own conclusions after reading my comments.  I cannot tell anyone else how to think.  I can be wrong, just as you can.

So – bearing this in mind, let’s take a look at the loudly-trumpeted ‘findings’ of the Faith and Sexuality Survey, as proclaimed by Bishop Bayes of Liverpool and Jayne Ozanne.

The Bishop of Liverpool states: ‘The statistics reflect lives which have been scarred and strained by mixed messaging of love, acceptance, condemnation and fear. … My own response begins in sorrow, and in repentance for what we have done to you, beloved children of God’.

Jayne Ozanne states: ‘The results provide strong evidence of the harm that attempts to change sexual orientation are reported to inflict’. 

Are these comments supported by the data?


They are based on assumptions that have been made about the data. The apparent ‘cause’ of behaviour or response x has been assumed to be the result of the respondent experiencing occurrence y.  But without cause-and-effect evidence, that is a logical fallacy.

‘My cat has four legs.  All cats have four legs.  Therefore all animals that have four legs are cats.  Therefore your dog is actually a cat.’

That’s a logical fallacy.  The ‘therefores’ are based on flawed reasoning, although on some level, some of them sometimes look as if they could be right.

Let’s look at some of the failures in reasoning exhibited in the report and the publicity around it.

Bishop Paul Bayes says in his Foreword to the report:

‘The findings of the report make hard reading for communities of faith, and in the case of my own Church, hard reading for an institution that believes it is built on the love of Christ’.

Flawed reasoning – the findings don’t prove that any harm has been caused by ‘the Church’.  The data describe people’s feelings about what they say has happened to them.  It’s not the same thing.  And why the Bishop bothers to mention ‘communities of faith’ when he clearly means ‘Christian churches’, I’m not sure.  If you were a Jewish, Muslim, Hindu or Sikh respondent, you or your background were, or could be, largely ignored after declaring your allegiance. So this was probably always about looking for problems with the Christian Church.

‘My own response begins in sorrow, and in repentance for what we have done to you, beloved children of God’.

Flawed reasoning – there is no evidence that ‘we’ have done anything, only people’s reported experiences, of which there is no adequate understanding at this stage.  There is a big jump between allegation and actual evidence. I’m not saying that respondents have lied, but we need to consider that most of the answers in the survey were subjective – meaning they were very open to personal variation when claiming the same experience. Is it appropriate to apologise for something that has not been proven?

The logical fallacy here is that we see an automatic assumption (a ‘therefore’) has been made:

Logical fallacy

All the problems cited in the responses are due to things that Christians have done to LGBT people. 

But as we shall see below, the responses that were provided are highly likely (as usual with online surveys) to have come from a sample of people who have been strongly self-selecting because they are already highly-motivated about LGBT grievances.  We know from much other research that this is typically what happens with online surveys – they tend to attract those with the strongest views on a subject.  This means, here, that the views expressed in the responses as a whole are not necessarily an accurate reflection of the experiences of the wider LGBT community who come to church or who have had a Christian background.

Despite this, Bishop Bayes has jumped straight in – with the best of intentions, no doubt – and assumed responsibility on behalf of the Church for causing hundreds of respondents to fall into mental ill-health, be abused and attempt suicide.  Is this a logical way to behave when the ink is barely dry?

‘You will see that far too many lives at far too young an age have been traumatised by a hidden inner conflict between their love of God and their innate desire to love another human being’.   

Flawed reasoning – see my last point.

‘The statistics reflect lives which have been scarred and strained by mixed messaging of love, acceptance, condemnation and fear. … More than half of those who had attempted to change their sexual orientation reported mental health issues …’

Flawed reasoning – putting aside the fact that ‘more than half’ does not mean ‘all’, the assumption is that the mental health issues, suicide attempts and suicidal thoughts reported in the survey are directly due to the teachings or actions of the Church, Christians and/or ‘religious leaders’.  In fact, mental health problems, suicide attempts and thoughts are highly associated with the raised rates of mental ill-health that we already know exist in among the wider LGBT community, whether individuals are religious or not; there is no evidence in the published survey results that reveals any cause-and-effect relationship between Church attendance, Christian faith or the actions of words of Christians and ‘harm’ resulting for LGBT people.

One logical conclusion to the comments of the bishop, above, could actually be to open the floodgates to legal claims for compensation for ‘injury’ as the bishop has already admitted liability.  Have we not learned anything from the apology to “Carol’ for Bishop Bell’s supposed and unproved ‘abuse’? Apparently not.

Jayne Ozanne states in her Director’s Report:

‘The results provide strong evidence of the harm that attempts to change sexual orientation are reported to inflict’.

Flawed reasoning – the data cannot attribute blame because the questions were not designed in such a way as to enable any link between experience and personal response at the time.  Some respondents may have attributed blame in their filled-in answers to the few questions that allowed ‘Other – please specify’ responses, but as those comments are not included in the published data, we don’t know.  If any of them related to non-Christian religious groups, why was this not stated?

Jayne Ozanne also itemised these examples as evidence of harm being inflicted on those who had experience of attempts to change their orientation:

  • More than half of those who had attempted to change their sexual orientation reported mental health issues and less than a third said that they “have gone on to lead a happy and fulfilled life”. Nearly half stated they had “found it hard to accept myself for who I am” and that they had “had to leave or change” their faith group.  

  • ‘Of those who suffered mental health issues (281 people), nearly a third (91 people) said they had attempted suicide while over two-thirds (193 people) said they had had suicidal thoughts. Two in five said they had self-harmed (significantly higher amongst women) and a quarter said they had suffered from eating disorders.

  • ‘43 people said they had been “given no choice and had to undergo it” and 22 people said they had undergone “forced sexual activity with someone of the opposite gender” in order to attempt to change’.

Flawed reasoning – see my comments above.  With regard to the ‘forced sexual activity’ reported, the answer is, on one level, straightforward –  these claims are describing criminal acts and ought to result in reporting to the police and probably prosecution. Nobody can be interested in calling that a ‘religious activity’. But apart from those claims, which we know no details of, there is no ‘strong evidence’ here, as Ms Ozanne claims.  There is a collection of people’s feelings about what has happened to them.  This does not mean that individuals have necessarily been subject to malicious treatment by Christians. Sadly, we already know that mental health problems are endemic in the LGBT community, whether individuals are religious or not.  But – apart from the forced sexual activity – there is absolutely no evidence here that any of this reported, and subjectively assessed, harm was caused intentionally -or even unintentionally – by Christians or the Church. Yet Ms Ozanne, like Bishop Bayes, is happy to claim by assumption and inference that all of the problems reported are due to Christians and the Church treating LGBT people badly.

All the ‘inflicted harm due to Christian activities’ trumpeted here is in the minds of the people analysing the data, because there is no designed link between experience A and experience B in the survey.  The notion of ‘harm’ carries with it the idea of A causing damage of some sort to B.  Yet B can also be damaged by his or her own responses, actions, thoughts, worries and so on, even without any input by A.  The Church and Christians may cause inadvertent harm to some LGBT people, just as adverts on the TV can, or comments on Twitter, but they cannot be blamed for all the mental health problems that LGBT people experience.  This survey simply can’t tell us.

The whole response of the leadership behind this survey has been an example of a logical fallacy being accepted as The Truth. Let’s look at the way the logical fallacy  behind all this flawed reasoning has developed from the survey data:

  • Some LGBT people, who may or may not be Christians, went to church and/or had a religious background of some sort (unspecified in any detail) when they were younger; and

  • some of them grew up with mental health problems and unhappy experiences; they also lived in the wider world and were subject to many secular pressures;

  • Therefore (that’s the logical fallacy bit) – we can say that all of their problems were caused by their religious experiences (we’ll ignore the secular pressures as we are only interested in evidence that religion Christianity is harmful)

  • Therefore (here we go again) – Christianity causes LGBT people to have mental health problems and damages their lives.

What about all the respondents who reported no religious background at all, but still reported later problems?  There were several hundred of them.  How do we explain their ‘issues’?  Were they only affected by secular pressures, and the church-raised LGBT people only affected by the Christian issues?

Is it valid to extrapolate the ‘assumed blame’ for all LGBT ills as reported onto the wider LGBT population?  Give it time – and a few more bishops jumping on the self-flagellation bandwagon – and it will.

I could go into more detail about various issues with the questions themselves, but for space and time reasons, I won’t.  There is one thing, however, that’s worth pointing out.   If respondents said they had experience of orientation change attempts, they got a whole lot more questions than the rest of us could see.

For example, Question 35 – ‘Why do you believe that it [sexual orientation change attempts] should be made illegal?’  Nothing wrong with that, of course, but why were we not all asked our opinion on that point?  Are our opinions on this invalid?  Are we not supposed to have an opinion on what should/not be made illegal in our country?

Why were all these extra questions not visible to the rest of us?  There could have been a ‘not applicable’ option.  Strangely, the figures don’t add up on these either.  If you look at the report and scroll down to Questions 25- 35, none of these appeared to respondents who said they had not experienced orientation attempts.  Yet when you look at the figures above each graph, the numbers clearly state that those who couldn’t even see the questions had ‘skipped’ them.  How can you skip a question you have never seen?  I have screen caps of all the questions that I was asked and completed.  These were not among them.  It’s worth noting as well that in the earlier questions, ‘skipped’ has been used to cover all those who did actually skip a viewed question, but also those who said ‘not applicable’.  They’re not the same thing.

This is very poor research practice.  In my opinion, it’s deceitful.  I’ll leave readers to reach their own conclusions about this.


Back in my December posts about this, I wrote about HARKing. This happens when exploratory research is wrongly used to make claims about it that can only be made of conclusive research (cause-and-effect research).  It’s called HARKing because it means ‘hypothesis after the results are known’ – which you can’t do based on the findings of exploratory research.  It’s also referred to as the ‘We knew it all along’ method of interpreting research findings.

I said at the time, ‘I will be very interested to see how the ‘findings’ of this research will be trumpeted in the media and at General Synod next year.  There is an illicit practice known in the research world as HARKing.  I suspect we are about to see a lot of it.’

So did we see any HARKing this week at General Synod?  Is the Pope Catholic? (Answers on a postcard, please).

Ms Ozanne obliged by saying, in her Director’s report:

 ‘For many, much of this document will confirm what they already know regarding the dangers of “conversion therapy” ‘.

And there it is.  One wonders why they even bothered to do the survey.

Definition issues

In my earlier posts I also made a point of saying how difficult it can be to address ‘conversion therapy’ due to the vague or very broad definitions which tend to be used by activists.

We can now see what the OF’s definition of conversion therapy includes, by looking at the response options provided in Question 28 in the survey.  This question asked what form of conversion therapy a respondent had undergone. The religious belief and/or practice response options were:

  • Private prayer
  • Prayer with close friends
  • Fasting
  • Deliverance ministry (unspecified in the survey, keeping the definition suitable flexible to include anything that could be counted as such), with friends, or more formally in a church setting
  • Emotional healing (unspecified, as the last point) with friends, or more formally in a church setting
  • Counselling with official religious ministry
  • ‘Other’ – which leaves the field wide open for any allegation of anything at all.

I don’t know about you, but as a practising Christian I find this very frightening.  Here we have a piece of poor research that is being trumpeted as evidence that Christian belief and practice causes harm, including suicide attempts, among LGBT people.  The Church of England appears to be swallowing all this whole.  The Bishop of Liverpool has admitted liability on behalf of the Church.

Given that the OF must, by default, accept these activities as forms of conversion therapy, and they are calling for such therapy to be banned, what does this say about the future for undertaking these things?  Private prayer?  Prayer with friends?  Fasting?  Healing and counselling?  Are we to take it that these activities are to be banned or controlled in some way? By whom?  How?

Starting to hear any alarm bells ringing?

The other definition issue that I raised in my earlier posts related to the identification of the ‘religious groups’ that Bishop Bayes referred to in his appeal to stop ‘conversion therapy’ in churches.  Well, as with conversion therapy, we now know the definition of ‘religious groups’.

Yes, dear reader, if you belong to the Church of England, or indeed, any other Christian church, it’s you.

All we need to do is look at the same response options in Question 28, plus the options in Question 21, which asked who had advised respondents to try to change their orientation.  We see that the following people must be included in ‘religious groups’ because they are cited as people who may cause harm by suggesting orientation change:

  • Friends, Religious
  • Leaders, Religious
  • Parent, Religious
  • Family member, Religious
  • Any religious person offering emotional healing, religious counselling or religious deliverance ministry.

Got that?  A ‘religious group’ can be a Christian praying with an LGBT friend.  That means a religious group can be a single Christian, as the LGBT person who has requested prayer does not need to be a believer.  While the survey refers repeatedly to ‘religious’ leaders, friends and so on, we can probably take it that they are really interested only in Christian groups or individuals.  Or are the findings going to be recommended to, and promoted as part of ‘interfaith dialogue’ with Jewish, Muslim, Hindu and Sikh leadership bodies as well as to the General Synod?  Are they?  Or do they not wish to cause offence, so maybe not?

Want to start thinking about your future ability and opportunities to engage in prayer with friend?  To share a fast with a friend who asks you to do so?  To offer or provide counselling according to Biblical principles to a friend if asked? To pray for healing if asked? Think again, folks!  Your time may be limited.  According to Bishop Bayes, all this is conversion therapy and must be banned from church.

The Church of England will now be on the path to accepting the definitions of ‘conversion therapy’ that the OF has laid out in its survey. In effect, this has been a clever way of getting a definition laid out through the side door. Is Our Great Leader Justin (peace be upon him) going to refuse to accept it?  Anyone want to open a book on that?

If you are not hearing those alarm bells yet, you should be.


We already know from a wide range of medical research that sadly, the LGBT community suffers far higher rates of mental illness than the wider population.  The Ozanne Foundation has used this survey to see if the Church and Christians (let’s not worry about the other major religions just now) are making this worse.  That’s what this survey is about:  does the Church make people’s mental health worse?  Does Christian, Bible-based belief make it worse?  We don’t know it from this survey, because this survey can’t tell us.

By claiming that it does, the Bishop of Liverpool and Jayne Ozanne are doing the LGBT community a disservice.  The evidence is not there.  All we have is a collection of responses about what people say they have experienced, with no digging deeper or trying to understand what it all means and why it may have happened.  There is nothing wrong with doing exploratory research like this.  It can be very useful for seeing what the scope of a perceived problem may be, who may be affected by it, or indeed, if there is a problem at all.  From such foundations, future – effective and well-designed – in-depth research with random sampling (as much as it is possible) may the tell us some actual truths.  What we cannot do is claim findings of cause-and-effect from exploratory research, yet that is what we are being asked to believe.

By proclaiming what may be a falsity – that the Church is responsible for mental ill-health and suicide/attempts, may only cause further anguish by ‘confirming’ what many LGBT people think they know.  This could actually add to people’s stresses.  But it may not be true on a large scale.  Likewise, further decent research may show that there is no measurable effect on the mental health of LGBT people caused by going to church or being a Christian.  This could actually be good for LGBT people because it could then allow them to focus on what other things may be causing their problems – which is what we all wish for them.

In order to get the sort of evidence needed to properly inform policy and practice, we need high-quality, robust, conclusive research.  I don’t this we’ll get it any time soon. Why not? Because to do such research requires qualified, experienced, independent, objective researchers to do it.  And they will have to point out to the research funders paying them (who are likely to be on the pro-LGBT side of the argument) that such research can sometimes produce evidence that the interested parties do not want to hear.  The problem may turn out to be a lot smaller or less serious than they thought.  It may turn out that the problem is not the ‘fault’ of the people or organisations that they want to blame.  Worst of all, it may show that the only way to resolve the problem is for the perceived ‘victims’ to change their own ideas rather than expecting everyone else to change.  This is true of any proposed cause-and-effect or evaluative research.

Concentrate on what you can control is my advice.  You cannot tell other people what to think. LGBT people cannot tell traditionalists what to think and feel and traditionalists cannot tell LGBT people what to think and feel.  This is a two-way street and not all the change can be on one side only.

Am I being pessimistic or biased here?  Well, there are plenty of research designers out there who are better at this than I am.  They will also be looking at this research and seeing its faults, even if only in private.  I’m not the only one who is looking at this research and thinking it’s weak.

Do good research and get good data that can really show what’s going on.  That is the best way to help those who need help. Pandering to people’s fears and biases works both ways – yes, the traditionalists may need to budge a bit.  But so do the LGBT activist community in church.  You can’t have everything you want. Get over it. What it all comes down to is this: do we believe the Bible’s teaching on human sexuality?  Yes or no?

Anyone want to bet on the outcomes of all this for the Church?  Do we now have any doubts at all as to the likely content and conclusions of the Living In Love and Faith report next year?  Just a thought!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s