Jonathan Hugh Williams
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Interests
Jonathan Hugh Williams has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
- Inga Williams
Department of Women’s and Reproductive Health, University of Oxford
1 shared publication - Julian Kenyon
The Dove Clinic for Integrated Medicine
1 shared publication - Karl J. Morten
The Nuffield Department of Women’s and Reproductive Health, The Women Centre, The University of Oxford, The John Radcliffe Hospital, Headington, Oxford, UK
1 shared publication
Latest Publications
- Open Access
- 17 September 2025
- Published by Center for Open Science in PsyArXiv
Sed quis custodet? Sentencing intentions of judges are more variable than those of lay people.
Background: There is a general assumption that judges assess crimes more reliably than jurors or the general public. Reliability is fundamental for trust and validity. Here, I compare the variability of sentencing intentions of judges and lay people.Quantifying the variability of sentencing is difficult, because sentencing guidelines impose boundaries on sentence lengths, that cause problems for estimation. I compared two ways to circumvent these problems: (1) by analysing actual sentences as proportions of the guideline range, or (2) by assuming that observed sentences reflect a latent distribution of “true” intentions. In both cases, the primary outcome was the variability of judges’ sentences compared with lay people’s.Methods: I analysed published sentencing intentions of professional judges and lay people for the same fictional crime. The source study’s sentencing guidelines allowed terms in the range 3-20 years, and over half the participants’ sentencing intentions were at these boundaries. I accounted for these boundary effects by transforming the raw sentence intentions to either (1) to proportions of the allowed range, or (2) to an ordinal variable, with 4 levels (0-9, 10-14, 15-19 or 20 years), representing a latent normal distribution. I then analysed the transformed data using distributional regressions, to assess ‘dispersion’ (reliability) in relation to rater status (judge, lay person)Results: Accounting for floor and ceiling effects importantly improved estimation of reliability. Judges’ sentencing intentions were more variable than those of lay people in both analyses. The proportional (relative) and ordinal (absolute) methods of analysis estimated lower or higher mean sentences for judges.Conclusion: Distributional regression can assess differential variability, as an indicator of reliability. The present results challenge the assumption that judges assess crimes reliably. This may help to explain the public’s lack of trust in the judicial system. It also throws into question the decisions of single expert raters in other high-stakes contexts.
- 17 September 2025
- Published by Center for Open Science in PsyArXiv
Open AccessSed quis custodet? Sentencing intentions of judges are more variable than those of lay people.
Background: There is a general assumption that judges assess crimes more reliably than jurors or the general public. Reliability is fundamental for trust and validity. Here, I compare the variability of sentencing intentions of judges and lay people.Quantifying the variability of sentencing is difficult, because sentencing guidelines impose boundaries on sentence lengths, that cause problems for estimation. I compared two ways to circumvent these problems: (1) by analysing actual sentences as proportions of the guideline range, or (2) by assuming that observed sentences reflect a latent distribution of “true” intentions. In both cases, the primary outcome was the variability of judges’ sentences compared with lay people’s.Methods: I analysed published sentencing intentions of professional judges and lay people for the same fictional crime. The source study’s sentencing guidelines allowed terms in the range 3-20 years, and over half the participants’ sentencing intentions were at these boundaries. I accounted for these boundary effects by transforming the raw sentence intentions to either (1) to proportions of the allowed range, or (2) to an ordinal variable, with 4 levels (0-9, 10-14, 15-19 or 20 years), representing a latent normal distribution. I then analysed the transformed data using distributional regressions, to assess ‘dispersion’ (reliability) in relation to rater status (judge, lay person)Results: Accounting for floor and ceiling effects importantly improved estimation of reliability. Judges’ sentencing intentions were more variable than those of lay people in both analyses. The proportional (relative) and ordinal (absolute) methods of analysis estimated lower or higher mean sentences for judges.Conclusion: Distributional regression can assess differential variability, as an indicator of reliability. The present results challenge the assumption that judges assess crimes reliably. This may help to explain the public’s lack of trust in the judicial system. It also throws into question the decisions of single expert raters in other high-stakes contexts.
- Open Access
Disentangling effects of ethnicity, deprivation, and payment source on obstetric outcomes in American primigravidae: A structural equation model of observational data
Background: Women from ethnic minorities have worse obstetric outcomes. Possible reasons for this are (1) social deprivation; (2) different standards of obstetric care; and (3) intrinsic ethnic differences. Here I aim to disentangle (1)-(3). Methods: I constructed two path models of causal links between parental ethnicity and obstetric outcomes. The first, ‘no-racism’, model estimated independent causal effects of ethnicity, deprivation and payment source on pregnancy and birth outcomes. The second ‘realistic’ model additionally tested how far deprivation and payment source may mediate effects of ethnicity. Analyses of the models used Bayesian estimation. I analysed both the full sample of complete data and a random 1% sample. Findings: Data were complete for 762786 births. The ‘no-racism’ model did not fit the data, but the ‘realistic’ model fitted adequately. It indicated that ethnicity, social deprivation, and private funding for care all adversely affected outcomes: (i) African American and Hispanic ethnicity caused deprivation; (ii) deprivation increased pregnancy hypertension, shortened gestation and reduced birthweight; (iii) private funding directly increased pregnancy hypertension and indirectly shortened gestation; (iv) participation in the Supplemental Nutrition Program for Women, Infants and Children (WIC) counteracted adverse effects of deprivation. (v) independently of (i)-(iv), ethnic-minority parents had shorter gestation and lighter babies. Interpretation: Deprivation largely accounts for adverse obstetric outcomes in ethnic minorities. Private funding may also worsen pregnancy hypertension, but WIC improved outcomes. The uniformity of adverse birth outcomes for all ethnic minorities suggests that these result from a common factor, which may be systemic racism. Policies to reduce deprivation and increase government-funded care could importantly improve obstetric outcomes, irrespective of ethnicity. Funding: none – I undertook the study at home. Research in Context: Evidence before this study: Many studies during the past century have shown that ethnic minorities have worse social deprivation and worse access to health services. Ethnicity, deprivation and care can all determine health outcomes, and ethnic-minority mothers have worse obstetric outcomes. However, the independent contributions of ethnicity, deprivation and care to these adverse outcomes are unknown. Added value of this study: I present here causal model of routine observational data that differentiates direct and indirect effects of ethnicity, deprivation and payment source on obstetric outcomes. The model allows (a) deprivation to mediate effects of ethnicity and (b) payment source to mediate effects of both ethnicity and deprivation. Hence, this model can disentangle the “intertwined” effects of ethnicity, deprivation and payment source on obstetric outcomes. The model also examines effects of participation in the Supplemental Nutrition Program for Women, Infants and Children on outcomes. The model fitted a 1% sample of the data after Bayesian estimation – so it bears interpretation as a representation of the real-world causal structure of the data. In the model, minority ethnicity causes deprivation and medical insurance and all of these factors independently determine adverse obstetric outcomes. Notably, medical insurance and private payment may increase the risk of pregnancy hypertension and consequently shorten gestation. Participation in WIC was beneficial. Implications of all the available evidence: Causal modelling of routine natality data may allow effective audit of health care in its social context. Understanding causes of poor outcomes can enable prediction of effects of policy change. The present results indicate that policies to ameliorate social deprivation and expand access to WIC and government-subsidised care should improve obstetric outcomes – with long-term benefits for both mothers and their babies. Extrapolating beyond obstetrics, the present results may help to illuminate mechanisms of the healthcare crisis in America.
Open AccessDisentangling effects of ethnicity, deprivation, and payment source on obstetric outcomes in American primigravidae: A structural equation model of observational data
Background: Women from ethnic minorities have worse obstetric outcomes. Possible reasons for this are (1) social deprivation; (2) different standards of obstetric care; and (3) intrinsic ethnic differences. Here I aim to disentangle (1)-(3). Methods: I constructed two path models of causal links between parental ethnicity and obstetric outcomes. The first, ‘no-racism’, model estimated independent causal effects of ethnicity, deprivation and payment source on pregnancy and birth outcomes. The second ‘realistic’ model additionally tested how far deprivation and payment source may mediate effects of ethnicity. Analyses of the models used Bayesian estimation. I analysed both the full sample of complete data and a random 1% sample. Findings: Data were complete for 762786 births. The ‘no-racism’ model did not fit the data, but the ‘realistic’ model fitted adequately. It indicated that ethnicity, social deprivation, and private funding for care all adversely affected outcomes: (i) African American and Hispanic ethnicity caused deprivation; (ii) deprivation increased pregnancy hypertension, shortened gestation and reduced birthweight; (iii) private funding directly increased pregnancy hypertension and indirectly shortened gestation; (iv) participation in the Supplemental Nutrition Program for Women, Infants and Children (WIC) counteracted adverse effects of deprivation. (v) independently of (i)-(iv), ethnic-minority parents had shorter gestation and lighter babies. Interpretation: Deprivation largely accounts for adverse obstetric outcomes in ethnic minorities. Private funding may also worsen pregnancy hypertension, but WIC improved outcomes. The uniformity of adverse birth outcomes for all ethnic minorities suggests that these result from a common factor, which may be systemic racism. Policies to reduce deprivation and increase government-funded care could importantly improve obstetric outcomes, irrespective of ethnicity. Funding: none – I undertook the study at home. Research in Context: Evidence before this study: Many studies during the past century have shown that ethnic minorities have worse social deprivation and worse access to health services. Ethnicity, deprivation and care can all determine health outcomes, and ethnic-minority mothers have worse obstetric outcomes. However, the independent contributions of ethnicity, deprivation and care to these adverse outcomes are unknown. Added value of this study: I present here causal model of routine observational data that differentiates direct and indirect effects of ethnicity, deprivation and payment source on obstetric outcomes. The model allows (a) deprivation to mediate effects of ethnicity and (b) payment source to mediate effects of both ethnicity and deprivation. Hence, this model can disentangle the “intertwined” effects of ethnicity, deprivation and payment source on obstetric outcomes. The model also examines effects of participation in the Supplemental Nutrition Program for Women, Infants and Children on outcomes. The model fitted a 1% sample of the data after Bayesian estimation – so it bears interpretation as a representation of the real-world causal structure of the data. In the model, minority ethnicity causes deprivation and medical insurance and all of these factors independently determine adverse obstetric outcomes. Notably, medical insurance and private payment may increase the risk of pregnancy hypertension and consequently shorten gestation. Participation in WIC was beneficial. Implications of all the available evidence: Causal modelling of routine natality data may allow effective audit of health care in its social context. Understanding causes of poor outcomes can enable prediction of effects of policy change. The present results indicate that policies to ameliorate social deprivation and expand access to WIC and government-subsidised care should improve obstetric outcomes – with long-term benefits for both mothers and their babies. Extrapolating beyond obstetrics, the present results may help to illuminate mechanisms of the healthcare crisis in America.
- Open Access
An approach to finding specific forms of dysbiosis that associate with different disorders
Background Many disorders display dysbiosis of the enteric microbiome, compared with healthy controls. Different disorders share a pattern of dysbiosis that may reflect ‘reverse causation’, due to non-specific effects of illness-in-general. Combining a range of disorders into an ‘aggregate non-healthy active control’ (ANHAC) group should highlight such non-specific dysbiosis. Differential dysbiosis between the ANHAC group and specific disorders may then reflect effects of treatment or bowel dysfunction, or may potentially be causal. Here, we illustrate this logic by testing if individual genera can differentiate an ANHAC group from two specific diagnostic groups. Methods We constructed an ANAHC group (n=17) that had 14 different disorders. We then used random forest analyses to test differential dysbiosis between the ANHAC group and two other disorders that have no known pathology, but: (i) symptoms of illness (Myalgic Encephalomyelitis / Chronic Fatigue Syndrome – ME/CFS – n = 38); or (ii) both illness and bowel dysfunction (ME/CFS comorbid with Irritable Bowel Syndrome – IBS – n=27). Results Many genera differentiated the ANHAC group from co-morbid IBS. However, only two genera - Roseburia and Dialister – discriminated the ANHAC group from ME/CFS. Conclusions Different disorders can associate with specific forms of dysbiosis, over-and-above non-specific effects of illness-in-general. Bowel dysfunction may contribute to dysbiosis in IBS via reverse causation. However, ME/CFS has symptoms of illness-in-general, but lacks known pathology or definitive treatment that could cause dysbiosis. Therefore, the specific dysbiosis in ME/CFS may be causal. [230 words] Contribution to the field Many disorders associate with enteric dysbiosis. The pattern of dysbiosis is largely consistent between unrelated disorders, which suggests that it mainly reflects non-specific secondary effects of illness-in-general (e.g. due to changes in activity levels, or diet). However, faecal microbiome transplantation (FMT) can be therapeutic in some disorders. This implies that unique features of dysbiosis may cause those specific disorders. Here, we propose a way to assess causal effects of dysbiosis, by testing if individual genera can discriminate individual disorders from an ‘aggregate non-healthy active control’ (ANHAC) group. Dysbiosis in the ANHAC group can control for non-specific effects of illness-in-general on the microbiome and so highlight potentially-causal forms of dysbiosis in specific disorders. This approach may provide insight into pathogenetic mechanisms of individual disorders and help to design specific forms of FMT to counteract them.
Open AccessAn approach to finding specific forms of dysbiosis that associate with different disorders
Background Many disorders display dysbiosis of the enteric microbiome, compared with healthy controls. Different disorders share a pattern of dysbiosis that may reflect ‘reverse causation’, due to non-specific effects of illness-in-general. Combining a range of disorders into an ‘aggregate non-healthy active control’ (ANHAC) group should highlight such non-specific dysbiosis. Differential dysbiosis between the ANHAC group and specific disorders may then reflect effects of treatment or bowel dysfunction, or may potentially be causal. Here, we illustrate this logic by testing if individual genera can differentiate an ANHAC group from two specific diagnostic groups. Methods We constructed an ANAHC group (n=17) that had 14 different disorders. We then used random forest analyses to test differential dysbiosis between the ANHAC group and two other disorders that have no known pathology, but: (i) symptoms of illness (Myalgic Encephalomyelitis / Chronic Fatigue Syndrome – ME/CFS – n = 38); or (ii) both illness and bowel dysfunction (ME/CFS comorbid with Irritable Bowel Syndrome – IBS – n=27). Results Many genera differentiated the ANHAC group from co-morbid IBS. However, only two genera - Roseburia and Dialister – discriminated the ANHAC group from ME/CFS. Conclusions Different disorders can associate with specific forms of dysbiosis, over-and-above non-specific effects of illness-in-general. Bowel dysfunction may contribute to dysbiosis in IBS via reverse causation. However, ME/CFS has symptoms of illness-in-general, but lacks known pathology or definitive treatment that could cause dysbiosis. Therefore, the specific dysbiosis in ME/CFS may be causal. [230 words] Contribution to the field Many disorders associate with enteric dysbiosis. The pattern of dysbiosis is largely consistent between unrelated disorders, which suggests that it mainly reflects non-specific secondary effects of illness-in-general (e.g. due to changes in activity levels, or diet). However, faecal microbiome transplantation (FMT) can be therapeutic in some disorders. This implies that unique features of dysbiosis may cause those specific disorders. Here, we propose a way to assess causal effects of dysbiosis, by testing if individual genera can discriminate individual disorders from an ‘aggregate non-healthy active control’ (ANHAC) group. Dysbiosis in the ANHAC group can control for non-specific effects of illness-in-general on the microbiome and so highlight potentially-causal forms of dysbiosis in specific disorders. This approach may provide insight into pathogenetic mechanisms of individual disorders and help to design specific forms of FMT to counteract them.