Guidance for reporting an a randomised trial of a social or psychological intervention

Using problem-solving therapy to reduce depressive symptom severity among older adult methadone clients: A randomized clinical trial

Impact of a social-emotional and character development program on school-level indicators of academic achievement, absenteeism, and disciplinary outcomes: A matched-pair, cluster-randomized, controlled trial.

For specific guidance see https://trialsjournal.biomedcentral.com/articles/10.1186/s13063-018-2735-z/tables/3.

Social and psychological intervention trials

National data suggest that 15% to 25% of women will be the victim of an attempted or completed rape during their lifetime. Research suggests college women are at greater risk for sexual victimization than women in the general population. The mental health consequences of sexual assault are serious. Women who are victims of sexual violence have higher and more severe rates of posttraumatic stress disorder (PTSD) than survivors of accidents and natural disasters. In addition to PTSD, there are many other insidious effects of sexual violence, which include psychological distress, physical distress, interpersonal problems, and increased risk for sexual revictimization.

Specific to cluster randomised trials

Sampling was based on a cluster randomized approach with schools, rather than individuals or classes, as the randomization units in order to minimize possible contamination or spillover effects between treatment conditions.

Specific objectives or hypotheses: social and psychological intervention trials

We thus hypothesize that sexual safety among HIV-positive men would be facilitated by self-efficacy and skills for enhancing social support and coping with HIV, modulating negative affect, enhancing HIV disclosure, and enhancing information and motivation specifically around sexuality…. We [also] hypothesized that overall unprotected anal intercourse [UAI] would lessen only moderately, whereas transmission risk—UAI that may transmit HIV to uninfected partners—would show significant intervention effects.

Specific objectives or hypotheses: Cluster randomised trials

The central question we addressed is, relative to the “business-as-usual” control condition, what is the effect of assignment to receive the Open Court Reading professional development and curricular materials on the spring literacy achievement outcomes of elementary school classrooms?

How the intervention is hypothesised to work: social and psychological intervention trials

Parent–infant interaction that is attuned and in which the parent is able to ‘read’ the child's communicative signals promotes positive social and communicative development in all children. Infants at-risk for autism often show ‘weak’ or distorted communicative signals which parents can struggle to recognise and respond to accurately. The iBASIS intervention was designed specifically to reverse such disrupted patterns of early parent–infant interaction, with the hypothesis that there would be consequent positive effects on other infant developmental markers and emerging prodromal autism symptoms…. iBASIS-Video Interaction for Promoting Positive Parenting (iBASIS-VIPP) uses video-feedback to help parents understand and adapt to their infant's individual communication style to promote optimal social and communicative development.

Example 1: social and psychological interventions

We employed a 2 (intervention: CBT (cognitive-behavioural therapy) vs. GHE (general health education)) x 3 (time: end of counseling at 2 weeks, follow-ups at 3 and 6 months) mixed factorial design. Participants … were randomly assigned to receive either CBT or GHE at a 1:1 ratio (n = 77 per intervention group).

Example 2: cluster randomised trials

We conducted a stratified randomized pretest–posttest controlled design study by enrolling passive recreation areas (PRAs) within public parks in 3 annual waves. After completion of the pretest assessment, parks were randomized by an independent biostatistician in an unequal 1:3 allocation ratio to treatment (shaded) versus control (unshaded) stratified by city, wave, and pretest use of the study PRA.

Originally, each teacher had three seventh-grade classes participating in the study. Each class was randomly assigned to receive one of the treatments so that each teacher taught all three treatments. The purpose of this restricted random assignment was to reduce the potential confounding effect of instructional methods. However, after the school year started, Teacher C lost one of her seventh-grade classes (the third treatment) because of a change of assignment to teach eighth-grade science.

Example 1: social and psychological interventions

Families were referred to the program by schools (30%), community-based agencies (22%), health care clinics (21%), self (16%), or public social services (12%). Participants were screened according to the criteria listed above and recruited during the 4-year period (1997–2001). Of the 302 families screened, 216 met the eligibility criteria…. The targeted families resided in Baltimore’s Westside Empowerment Zone (i.e., federally designated as an area of extreme poverty, unemployment, and general economic distress) and had at least one child between the ages of 5 and 11. Eligibility included (a) a concern by the referring person that at least 1 of 19 neglect subtypes (e.g., unsafe housing conditions, inadequate supervision, inadequate/delayed health care) was occurring at a low level but not at a level that Child Protective Services (CPS) would accept for investigation; (b) at least two additional risk factors for neglect related to the child (e.g., behavioral problem; physical, developmental, or learning disability; more than three children) or the caregiver/family (e.g., unemployment/ overemployment, mental health problem, drug or alcohol problem, domestic violence, homelessness); (c) no current CPS involvement; and (d) caregiver expressed willingness to participate in the FC program.

Example 2: cluster randomised trials

Eligible schools were officially registered, had at least four classrooms and 120 students, were located in close proximity to other schools (i.e., ∼10 km or one hour walking), were in a secure zone at the time of school recruitment (e.g., no movement of armed groups), accessible by motorbike, and presumably not receiving support similar to OPEQ by other private, local, or international agencies.

Example 3: Eligibility criteria for settings and those delivering the interventions

The following study inclusion criteria were used to control for any further variability in classroom characteristics across program type: (1) teachers with a bachelor’s degree or an associate’s degree and working towards a bachelor’s; (2) programs with moderate to high quality as measured by the NC (North Carolina) star-rating system (3–5 stars out of 5 stars total), (3) use of the Creative Curriculum (a state-approved MAF (More at Four Pre-Kindergarten Program) curriculum and the predominant curriculum used by MAF classrooms), (4) classroom enrollment of at least four Latino ELL (English Language Learner) children, but not to exceed 85% of total enrollment, and (5) use of English as the primary language of instruction.

Our study site is St. Louis County, Missouri. We chose St. Louis County because it is part of a large metropolitan area with significant crime problems. In areas of the county patrolled by St Louis County Police Department (SLCPD), the 2012 violent crime rate was 244.6 per 100,000 while the property crime rate was 2063.8. Covering more than 500 square miles, with over 1 million residents, the county is the 34th largest in the U.S. and contains 17% of the state’s population, although the SLCPD provides primary police services to just over 400,000, including to more than 90 municipalities that contract for services. The SLCPD employs just over 800 sworn and 240 civilian personnel and is an internationally accredited, full-service department. Officers have fairly stable geographic assignments, ensuring some continuity across the treatment period.

Example 1: social and psychological interventions

The integrated 12-Step facilitation (iTSF) intervention consisted of one 60–75-minute individual session, followed by eight weekly, 90-minute iTSF group sessions (n = 2-5 per group). Immediately prior to the fifth group therapy session, participants completed a second 30–50-minute individual ‘booster’ session. iTSF employed a ‘Socratic’ therapeutic questioning style to promote adolescent attention, verbal engagement and participation in discussion of topics, and therapists also used a variety of TSF strategies. At the end of each group, participants identified a sober activity goal for the week and reported on it at the beginning of the session the following week. Six sessions were based around a recovery-related topic, and two sessions invited members of 12-Step organizations to share their recovery story and experience. One therapist had a master’s degree in social work and was a certified Licensed Alcohol and Drug Counselor (LADC)-I with more than 5 years of experience. The other was in a clinical psychology doctoral training program, with several years of supervised clinical experience in substance use disorder (SUD) and mental health treatment. Both therapists had specific experience in cognitive behavioral therapy, 12-Step philosophy and principles and group-based interventions for individuals with SUD particularly with adolescents. Additionally, prior to beginning the study, therapists each attended five 12-Step meetings to familiarize themselves with the meeting format.

Example 2: cluster randomised trials

In the parent and youth (PY) group, parents received the Familias: Preparando la Nueva Generación (FPNG) parenting curriculum, and youth received the youth-centered substance-use prevention program, keepin’ it REAL (kiR). In the youth group, youth received kiR, and parents did not receive any curriculum. In the control group, parents and youth received treatment-as-usual with respect to curricula offered at the schools.

Treatment adherence and competence were monitored through weekly supervision and supervisor review of audio-recorded sessions. … In addition, four iTSF sessions (three group and one individual) were selected at random and rated by two independent doctoral level clinicians on three dimensions: adherence to protocol, skill level and frequency and extensiveness of skills used. On average, adherence was rated at 96.4%, and skill level as 6.4 – where 6 = ‘very good’ and 7 = ‘excellent’. Frequency and extensiveness was rated as 3.7 – where 3 = ‘adequately’ and 4 = ‘extensively’…. Of 59 participants who completed a baseline assessment, four never received any treatment, two of whom actively withdrew and two of whom were unable to be contacted; by the mid-treatment assessment, three more individuals withdrew consent; by end of treatment, two more participants withdrew consent. The study took place from July 2013 to October 2015.

For more details, see: Kelly J. F., Yeterian J. D., Cristello J. C., Kaminer Y., Kahler C., Timko C. Developing and testing twelve-step facilitation for adolescents with substance use disorder: manual development and preliminary outcomes. Subst Abuse 2016; 10: 55–64.

The two study therapists treated patients in both treatment conditions to avoid the problem of differential therapist effects. Individual motivational enhancement therapy (MET) sessions were divided equally across the two therapists and therapists co-led all groups in both conditions.

Example 1: social and psychological interventions

Irritable bowel syndrome symptom severity was measured using the Irritable Bowel Syndrome Severity Scoring System, which measures the severity of pain, distension, bowel dysfunction and quality of life/global well-being. Assessments were administered at four time-points: baseline (pretreatment), post-treatment (2 months) and at 3 and 6 months post-treatment. A decrease of 50 points on this scale has been identified as a clinically significant change in symptom severity.

Example 2: cluster randomised trials

Viewsheds incorporate GIS methods to digitize the actual line-of-sight of CCTV cameras, which more accurately reflects camera coverage than traditional units of analysis, such as aggregate geographies (i.e., neighborhoods or police beats) and circular buffers drawn around camera sites. Researchers viewed the live feeds of all CCTV cameras in Newark and digitized the viewshed of each site within a GIS, … [which] created 75 separate CCTV schemes from the 146 individual viewsheds…. Catchment zones were created for each of the CCTV schemes [and served as the unit of measurement].

We attempted to obtain collateral reports on the participants’ alcohol and drug use at the 12 month follow-up…. Because of the relatively low rate of apparent under reporting by participants relative to collaterals, and the small percentage of participants for whom collateral data were available, participant self-reports were used in the analyses of alcohol use.

Example 1: social and psychological interventions

Power calculations were performed using SAS PROC POWER for the primary aims of evaluating the equivalence of the UP and SDPs and evaluating the efficacy of the UP and SDPs relative to a benchmark WLC and were based on conventional target values of power = 0.80 and α = .05. With an allocation ratio of 2:1 for active treatment to WLC groups, results of the power calculations indicated that a sample size of 91 individuals per active treatment group provided adequate power for the analyses of both equivalence and superiority…. The equivalence margin of 0.75 Anxiety Disorders Interview Schedule clinical severity rating (ADIS CSR) units was selected based on available meta-analytic reviews of cognitive behavioral therapy outcome studies and recommendations for selecting a priori equivalence limits.

Example 2: cluster randomised trials

For a medium effect size (d = 0.5, power = 0.80, α = .05), a sample size of 64 per group was needed to test for differences in means between two groups. For cluster RCTs, the sample size was adjusted according to the design effect (design effect = 1 + [cluster size −1] × ICC). As the design effect was 1.23 for parent and child outcomes and .277 for teacher reports, the required sample sizes were 158 and 433, respectively.

An independent data monitoring committee reviewed unblinded data for safety after the first 1,000 women in the study had given birth. In response to a lower than anticipated attrition rate, we stopped recruitment when 1,748 had been randomly assigned.

Random allocation was managed by the study statistician using computer-generated random numbers.

Example 1: social and psychological interventions

A blocked randomization scheme (blocks of 30) was used to yield balanced allocation of participants to treatment groups. Using a computerized adaptive minimization procedure, subjects were matched on suicide attempts or nonsuicidal self-injuries; psychiatric hospitalizations; history of suicide attempts and/or nonsuicidal self-injury; age; and a negative prognostic indicator of depression or a comorbid condition.

Example 2: cluster randomised trials

School Report Card data were used to stratify schools into strata ranked on an index based on [individual] demographic variables; (b) characteristics of the student populations; and (c) indicators of student behavior and performance outcomes. Schools were matched on index score, resulting in 19 strata. Matched pairs were randomly selected from within strata, with one school of each pair randomly assigned to either intervention or control.

Example 1: social and psychological interventions

The randomization list was transferred to a sequence of brown envelopes by writing the sequence of treatment names on the inside of the envelopes, which were then sealed. The sequence of envelopes was then ‘cut’ by taking approximately the first half of the envelopes and placing them at the end of the sequence so that no person involved in the trial would know the starting point of the randomization sequence and to preserve allocation concealment. The envelopes were then numbered.

Example 2: cluster randomised trials

Schools agreeing to participate were stratified by percentage of children receiving free school meals (dichotomised at the median) and were randomly allocated within stratum to treatment arm. One researcher generated the allocation schedule using the Stats Direct computer program, the research unit co-coordinator allocated the schools to treatment arm blind to the identity of each school, and a second researcher enrolled schools.

Example 1: social and psychological interventions

A study nurse telephoned a person at a randomization center who did not know the identities of the potential couples. The study nurse read the names from a list in order in which they had been assessed. Couples were randomly allocated by means of computer-generated random numbers. Every randomization result appeared in the program after the participants name was written and the person executing the randomization confirmed the process with her initials. This ensured that neither the study nurse, nor person doing the randomization could influence the result.

Example 2: cluster randomised trials

General practices were the unit of randomization and determined the patients’ group status. GPs were randomised by coin toss after they gave their written informed consent. 136 GPs (15.9%) gave written informed consent to participate and agreed to adhere to the DelpHi trial protocol…. GPs assessed the eligibility of patients (≥70 years, living at home) and systematically screened patients who met the inclusion criteria…. All persons eligible for the study will be screened for cognitive impairment. People who met the inclusion criteria and provided their written informed consent to participate were included.

Although participants and clinicians delivering the treatment could not be blinded to treatment assignment, assessors and clinicians conducting outcome assessments were blinded. In addition, participants were instructed at their follow-up assessment interviews not to reveal their treatment assignment.

The attention control condition was led by the same interventionists who led the hypnosis intervention sessions. However, the interventionists did not lead the attention control patients in imagery, relaxation, or even simple discussion. Rather the interventionists allowed patients to direct the flow of the conversation and provided supportive and empathic comments according to standardized procedures.

Example 1: social and psychological interventions

Linear random-effects models (hierarchical regression models) were implemented with random intercepts and slopes. These models estimate main effects for change from baseline to each assessment at 6, 12, and 18 months, main effect for the treatment, and interactions between the visit and treatment indicator variables. For each of the primary and secondary outcomes, separate intent-to-treat tests and estimates (with 95% CIs) of randomized group contrasts at 6, 12, and 18 months were obtained from the estimates of the respective time × treatment interactions. Potential confounding variables were evaluated by assessing whether baseline factors imbalanced between the treatment groups were related to outcome. Age was found to be a confounding variable and was controlled in all analyses.

Example 2: cluster randomised trials

Because schools were matched into pairs prior to randomization, the data presented here are nested: Children are nested in schools, and schools are nested in their matched pairs. To accommodate these design features, we calculated estimates of intervention impact on change in the primary child outcomes from preintervention baseline (fall 2004, Wave 1) to the first follow-up (spring 2005, Wave 2) using a series of two-level hierarchical linear models with random effects in HLM 6.02. In these models, Level 1 (child) included the preintervention baseline of the dependent variable and the child-level baseline covariates, and Level 2 (school) included a dummy variable indicating intervention condition as well as eight dummy variables representing the school pair matches.

Example 3: How missing data were handled

For continuous outcomes, we used the multiple imputation procedure from SAS; because the data were non-monotonically missing, we used the Markov Chain Monte Carlo procedure. We used all available data regarding demographic status, psychosocial variables, unprotected anal intercourse, and transmission risk partners to impute missing values on risk outcomes. Missing data correction for binary measures used the previous wave value.

Example 1

We examined effect moderators using multiple regression. In Step 1, baseline conduct problem score was entered, followed by intervention status and moderator variable. In Step 2, the interaction term (Potential Moderator x Intervention Status) was introduced. We examined mediators by assessing associations between change in putative mediator, change in outcome, and intervention status; conducting hierarchical multiple regressions; and assessing significance using the Sobel test.

Example 2

The process evaluation was based on an inductive thematic analysis. Recordings were transcribed, coded, and analyzed by two researchers. Each researcher drew on the other to assess rater reliability and interpretation. Each session observed was reported separately first using a grounded theory approach with no observational protocol. Themes emerged through considering sessions and of archival materials.

See Figure 1 https://trialsjournal.biomedcentral.com/articles/10.1186/s13063-018-2735-z/figures/1.

Two of the 502 BMI (brief motivational intervention) participants were administratively dropped from the study when it was discovered after randomization that they began working at the university survey research center collecting data for the study.

Recruitment occurred from April 2006 to January 2008 using radio, web-based, and newspaper advertisements for a smoking cessation intervention consisting of group therapy plus nicotine patch…. Smoking was assessed at 1 week-, 4 weeks- (end of behavioral treatment), 16 weeks-, and 26 weeks-post assigned quit date.

Recruitment was stopped once the study sample size was achieved because of study timelines and budget constraints, and the study was stopped once the 1-year follow-up assessments were completed.

Example 1

See Table 6 for number of participants for each analysis. The primary analyses used all available follow-up data and compared participants in their randomized groups, irrespective of the intervention they received. The sensitivity of the primary analyses was assessed including baseline school attendance, using a per protocol analysis (excluding three participants in the psycho-education group, two of whom did not fulfil criteria for CFS (chronic fatigue syndrome) and one who received 13 sessions of CBT (cognitive behavioural therapy)) and multiple imputation as an alternative method for handling missing data.

Example 2

See Table 6. Importantly for our analyses, allocation was not associated with attrition. Results from complete case analyses (CCA; not tabled) were also carried out and did not differ markedly from those reported here. All models were carried out on the intent-to-treat basis and estimated controlling for student sex and baseline values of the evaluated outcome.

Example 1: social and psychological interventions

See Table 2 for adjusted and unadjusted summary results for each study group and the estimated effect size on continuous and dichotomous outcomes.

Example 2: cluster randomised trials

See Table 2 for unadjusted summary results for each study group and the estimated effect size for each outcome. The average cluster size (number of participants in each preschool) was eight and the intracluster correlation (ρ) was 0.07. The design effect was thus 1.49.

Example 3: Availability of trial data

Data Availability: All relevant data are within the paper and its Supporting Information files.

Based on observed data, 24.3% (N=27/111) of the patients in the CBT (cognitive behavioural therapy) condition and 21.3% (N=26/122) in the psycho- dynamic therapy condition met the remission criterion at the posttreatment assessment…. At the posttreatment assessment, the odds ratio was 0.82 (95% CI=0.45–1.50), indicating that remission rates did not differ significantly.

Example 1

There were no significant moderator effects for single parenthood, very low income, teen parenthood, and baseline level of observed child deviant behavior. Child gender, depression, and age were significant moderators, interacting with intervention status to predict conduct problem outcome. The intervention produced better conduct problems for boys, children of more depressed mothers, and younger children. Mediator analyses found change in positive parenting skill predicted change in conduct problems.

Example 2

There was an openness and willingness from police officers to talk about their work in an accessible, engaging way. Young people were generally very well behaved and engaged in the sessions. Some facilitators were adept at holding information generated by the group and returning to it at relevant later points. The best facilitators did so in neutral, non-judgmental ways. However, balancing the quantity of material with the quality of interaction was the biggest practical challenge observed. Supporting materials were difficult to implement systematically or consistently.

Regarding major incidents, two suicide attempts, two accidental deaths related to psychotic symptoms, a serious fight where the patients sustained serious injuries and three patients initiating substance abuse were recorded in the control group. In the family intervention group minor incidents were detected such as starting sexual relationships with the risk of HIV infection, vagrancy, bouts of alcohol consumption, and aggressivity.

As described previously, statistical conclusion validity in the argumentation analysis was limited by not having a specific argumentation pretest covariate in the model, and instead a knowledge/reasoning pretest value for each student was used because a correlation was expected. Our claims about retention are similarly tempered (no argumentation pretest or no knowledge/reasoning retention measure). Other limitations of this study include the small sample size (58 students) and the short length of the intervention (10 hours of instruction, 4 hours of testing), yet the fact that we found significant and consistent differences despite these limitations speaks to the strength of the effect. Despite the teacher in this study having many years experience teaching both traditional and inquiry-based materials, he is undoubtedly more of an advocate of an inquiry-based approach. However, we believe the benefits of controlling variables by having the same teacher in both sections outweighed the potential bias created by a teacher being more comfortable in one approach than the other, and findings such as the comparable levels of student engagement shown in Table 4 suggest that the treatments were not strongly teacher-biased.

Example 1: social and psychological interventions

Although the ethnic makeup of the sample roughly approximated that of the county in which our research center was situated (Middlesex County, Massachusetts), the sample included fewer African Americans, more biracial children, and participants with higher parental education and socioeconomic status than the general Middlesex County community. Further studies are needed to examine the protocol’s efficacy in samples with greater socioeconomic and ethnic diversity, as well its effectiveness in community mental health settings. In addition, our criteria allowed for the exclusion of children judged too uncooperative or distractible to take part in the treatment (two children) or children deemed too clinically severe to wait 6 months to receive treatment, based on severe mood disorder, severe social isolation, severe impairment in school function or attendance, or severe OCD (obsessive compulsive disorder) (a total of seven children). These criteria generally excluded children who clinically would not be administered CBT (cognitive behavioral therapy) for anxiety disorders as their first treatment (i.e., they might be offered such treatment after their other symptoms were addressed). Therefore, study results can be generalized only to children whose anxiety disorders are not so severe as to cause school refusal or severe social isolation. In other regards, however, the sample appeared representative of clinical samples, with high comorbidity of anxiety disorders and with 69% in the borderline or clinical range on the CBCL (child behavior checklist) Internalizing scale. In addition, our extensive intake assessment battery, which required a total of four parent and/or child visits prior to randomization, deterred as many as 1 in 4 potential participants and may have selected for families who were especially motivated to take part in treatment.

Example 2: cluster randomised trials

In contrast to the lack of change shown in adolescents receiving creative workshops in Uganda, this intervention [in Indonesia], which includes structured creative activities as well as trauma-focused activities, did show effects on psychosocial well-being. It could therefore be considered a preliminary argument that increased structured interventions, which include trauma-focused activities, more effectively target PTSD symptoms. Corroboration of this argument can be found in the high-effect sizes of group CBT implemented in violence-affected schools in Los Angeles. However, previously mentioned qualitative research has shown the importance of addressing wider social problems caused by war, rather than purely focusing on PTSD complaints. In addition, specialized mental health professionals to implement CBT are usually unavailable in low-income settings. To resolve this tension, we propose that in complex emergencies, interventionists use a public health framework to tailor interventions to an appropriate population and referral level, based on investigated local needs, severity of complaints, available resources, and feasible and cost-effective interventions, while recognizing the importance of the social-ecological context. On the basis of these findings, the classroom-based intervention then qualifies as an appropriate intervention to target larger groups of children (especially girls) at risk, when stress-related symptoms are relevant.

The results from this trial differed from previous CBT (cognitive behavioural therapy) trials in two key areas. Only one patient (3%) did not complete the treatment. Previous IBS (irritable bowel syndrome) studies suggest that drop-out rates from CBT can be as high 40%. This may be because traditional CBT requires a substantial time commitment from patients. The most common reasons for dropping out are being unable to take time off work or childcare commitments. Having fewer sessions and sessions on the telephone may make the therapy more widely available. In addition, presenting treatment as self-management of a chronic condition rather than as a psychological therapy may be more acceptable to IBS patients. The treatment effects for symptom severity in this study are larger than those reported in many other CBT trials. This may be because of differences in the patient cohorts. As our study did not rely on GP (general practitioner) referral we may have accessed a cohort that seldom gets offered therapeutic intervention or perhaps even gets diagnosed. This is important, as our results suggest that treatment effects may be greater if patients are less disabled by their symptoms and less depressed. There is certainly evidence that depression in IBS is related to poorer treatment outcome. This study indicates that early intervention and diagnosis may not only make treatment more effective but also prevent the illness becoming more chronic and refractory to treatment.

We registered the study with the American Economic Association (AEARCTR-0000742).

The full trial protocol is available in the Supplement: https://jamanetwork.com/data/Journals/PSYCH/935708/YOI160049supp1_prod.pdf.

Example 1

This work was supported by a research grant from the National Institute on Drug Abuse (R01 DA015183-05) with cofunding from the National Cancer Institute, the National Institute of Child Health and Human Development, the National Institute of Mental Health, and the Center for Substance Abuse Prevention. The funders had no role in the design, conduct, analysis and reporting of the trial.

Example 2

The contributions of GP in reviewing the creative process instrument a few times are gratefully acknowledged. GP authored a few books such as Evidence based teaching—A practical approach (2006), Teaching today: A practical guide (2004) and How to be better at creativity (1996). He works for the Learning Skills Development Agency as a consultant on their Raising Quality and Achievement programme, assisting Action Research Development Projects in colleges, and assisting the Quality Improvement Team. He is a visiting examiner for the Institute of Education at London University. His experience includes physics teaching, managing teacher training, Inclusive Learning Facilitator, being a staff development officer, and managing college lesson observation.

The evaluation was conducted by the same team of people that designed the intervention. In order to promote objectivity of the evaluation we submitted the trial protocol for publication before the study began and submitted statistical analysis plans with prespecified primary outcomes before data were inspected, and these plans were scrutinized by an independent data-monitoring committee. As with all new projects, it is likely that the intervention was managed and implemented with greater expertise and enthusiasm by the project team than could be expected in subsequent iterations and scaling up of the intervention.

The SKCHH [Seattle–King County Healthy Homes] project was designed as a community-based participatory research project with overall sponsorship by Seattle Partners for Healthy Communities, an Urban Research Center funded by the U.S. Centers for Disease Control and Prevention. Seattle Partners is a multidisciplinary partnership of community agencies, community activists, public health professionals, academics, and health providers that supports community-based participatory research addressing social determinants of health…. Both the Seattle Partners Board and the steering committee sought to assure that the project benefited all participants. This led to the staggered intervention design with low- and high-intensity groups. This design assured that low-intensity group participants initially received some immediate benefit (including interventions known to be useful, such as bedding encasements) while ultimately receiving all the benefits accorded the high-intensity group. While this design may have reduced the study’s power to demonstrate an effect of the high- intensity intervention relative to a “pure” control group receiving no intervention, we felt such a design was not ethical.

As part of an incentive package to each participating school, all teachers received free OCR program materials and professional development supports throughout the year. Control schools from five districts received a core math program (Everyday Mathematics) and related professional development at no charge, whereas two districts received a cash incentive of $5,000 per school in each year of the study. In this sense, the treatment and control schools received a similar amount of resources, with treatment schools receiving additional supports for their literacy programs and control schools receiving additional supports for their core math or other educational programs. Additionally, the study provided incentives to participants. Teachers received $15 for completing a survey in the fall and spring of each year. In addition, teachers received up to $45 for completing an interview and for allowing researchers to observe their classrooms up to three times per year. A staff person from each school volunteered as a school liaison and received a $500 stipend per year for coordinating research activities at each school.