eprintid: 10158003
rev_number: 8
eprint_status: archive
userid: 699
dir: disk0/10/15/80/03
datestamp: 2023-05-11 15:56:15
lastmod: 2023-05-11 15:56:15
status_changed: 2023-05-11 15:56:15
type: conference_item
metadata_visibility: show
sword_depositor: 699
creators_name: O'Mara-Eves, Alison
creators_name: Kneale, Dylan
creators_name: Candy, Bridget
creators_name: Sutcliffe, Katy
creators_name: Oliver, Sandy
creators_name: Cain, Lizzie
creators_name: Hutchinson Pascal, Niccola
creators_name: Thomas, James
title: Recalibration as an approach to context in statistical meta-analyses
ispublished: pub
divisions: B14
divisions: J81
divisions: B16
divisions: UCL
keywords: Statistical meta-analysis, evidence synthesis, systematic review, context, generaliability, secondary data sources
abstract: Background: Using evidence from meta-analyses for local decision-making is hampered by the lack of explicit connection between the contexts in which interventions were evaluated and the context in which the evidence is to be applied. New methods are needed to address the question, ‘is there any evidence that the intervention will work differently in an area like mine?’ /

Methods: We reanalyzed a meta-analysis of school-based interventions to reduce fat intake. Using observational data, effect sizes were reweighted to reflect the contextual similarity of studies to four UK Local Authorities. Studies that were more like Local Authorities contributed more towards the pooled effect size. /

Results: The original meta-analysis found that the intervention was not effective in reducing fat intake. In the recalibration analyses, because contextually relevant studies showed greater effects, the recalibrated pooled effect sizes indicated a larger effect with a narrower confidence interval in three Local Authorities, but the interpretation remained unchanged in a fourth. The findings held under both fixed effect and random effects specifications. /

Discussion: For three of the local authorities examined, we have greater confidence that school-based interventions will have an impact on reducing fat intake. Recalibration may give a contextually-nuanced estimate for a given local context.
date: 2022-10-18
date_type: published
official_url: https://wwgs2022.mailchimpsites.com/
oa_status: green
full_text_type: other
language: eng
primo: open
primo_central: open_green
verified: verified_manual
elements_id: 1984257
lyricists_name: O'Mara-Eves, Alison
lyricists_id: AJOMA15
actors_name: O'Mara-Eves, Alison
actors_id: AJOMA15
actors_role: owner
full_text_status: public
pres_type: presentation
event_title: 2022 What Works Global Summit (WWGS)
event_location: Virtual conference
event_dates: 18 - 20 October 2022
citation:        O'Mara-Eves, Alison;    Kneale, Dylan;    Candy, Bridget;    Sutcliffe, Katy;    Oliver, Sandy;    Cain, Lizzie;    Hutchinson Pascal, Niccola;           O'Mara-Eves, Alison;  Kneale, Dylan;  Candy, Bridget;  Sutcliffe, Katy;  Oliver, Sandy;  Cain, Lizzie;  Hutchinson Pascal, Niccola;  Thomas, James;   - view fewer <#>    (2022)    Recalibration as an approach to context in statistical meta-analyses.                   Presented at: 2022 What Works Global Summit (WWGS), Virtual conference.       Green open access   
 
document_url: https://discovery.ucl.ac.uk/id/eprint/10158003/1/221018%20WWGS%20Recalibration%20as%20an%20Approach%20to%20Context%20in%20Statistical%20Meta-analyses_final.pdf