RESEARCH DISCIPLINE
Statistical Analysis for Clinical Psychology Research
Consulting for Treatment & Outcome Studies
Clinical psychology statistical analysis for therapy trials, pre-post evaluations and symptom-scale studies, with dropout, meaningful change and unstable estimates in small samples addressed, and the results prepared to the reporting standard your journal expects.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
Statistical Analysis for Clinical Psychology Research
Clinical psychology statistical analysis is the statistical design, analysis and reporting of studies of psychological disorders and their treatment, such as therapy trials, pre-post evaluations and symptom-scale studies, most of which follow patients recruited in clinics through repeated symptom scores. The samples are small, patients leave treatment for reasons tied to how they are doing, and the candidate predictors and mediators often outnumber what the sample can support, so a workflow of separate time-point tests on those who completed treatment risks biased and unstable estimates. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, provides clinical psychology data analysis through in-house PhD-level statisticians, who agree the estimand, the handling of dropout and the number of parameters with you before anything is fitted.
WHEN TO ENGAGE
Statistical Support at Different Stages of Clinical Psychology Research
Researchers usually contact us at a recognizable point in a treatment study, from the first design questions to a reviewer's objection to the way dropout was handled.
Planning a Treatment Trial You are designing a therapy trial or pre-post evaluation and need the primary outcome, the sample your clinic can recruit and a plan for dropout fixed before the first patient is enrolled.
Analyzing Symptom Scores Over Time Your repeated symptom measures are collected, some visits are missing, and the model, the handling of dropout and the criterion for meaningful change must be chosen, run and checked.
Preparing a Clinical Manuscript Results need intervals, effect sizes, an account of participant flow and a methods paragraph prepared to CONSORT 2025 or to the APA reporting standards, as your target journal expects.
Reviewing a Completed Analysis A reviewer or advisor has questioned a completers-only analysis, many separate time-point tests or a mediation model fitted to a small sample, and the analysis needs an independent check.
RESEARCH CONTEXT
Research Designs and Data We Work With in Clinical Psychology
These are the datasets of clinical samples, symptom scales and treatment outcome studies. Counseling process and service-based wellbeing studies are described on the Counseling & Mental Health Psychology page.
01
Randomized Treatment Trials
Therapy compared with a waitlist, usual care or an active control, scored at baseline, after treatment and at follow-up, with patients treated by a few therapists.
02
Pre-Post and Case Series Studies
Outcome studies from one clinic or a single series of patients, often without a comparison group, where improvement must be separated from natural recovery.
03
Symptom Scale Data
Self-report and clinician-rated measures of depression, anxiety, trauma or related conditions, scored to the manual and checked for reliability in your sample.
04
Longitudinal Patient Cohorts
Patients followed through treatment and afterward, with symptom trajectories, remission and relapse, and attrition that differs between those doing well and badly.
05
Process and Mediator Studies
Data collected to test how a treatment works, such as alliance or thought patterns measured during therapy, usually in samples of modest size.
06
Diary and Experience-Sampling Data
Several short assessments a day over weeks, with unequal spacing and missed prompts, used to follow symptoms as they change within a person.
METHODOLOGICAL CONSIDERATIONS
Statistical Issues We Consider in Clinical Psychology Research
These are the checks made in treatment outcome statistics before any result is trusted, because small samples and informative dropout can change an estimate more than the choice of test.
Dropout and the Estimand Patients who leave treatment often differ from completers, so the estimand is stated, every randomized patient is analyzed with all available visits, and sensitivity analyses probe the missing-data assumptions.
Meaningful Change for Each Patient A symptom score is judged against its reliability, so that a patient's change is classed as reliable improvement, no change or deterioration, and the proportion in each class is reported beside the mean change.
Therapist Effects Patients of one therapist share an influence on outcome, so therapist is modeled where several are involved. With very few therapists that variance is estimated imprecisely, and the limit is stated.
Few Patients, Many Predictors A small sample supports few parameters, so the number of predictors is limited and fixed in advance, and estimates are shown with intervals, not chosen from many fitted models.
Timing of a Mediator A mediator measured at the same time as the outcome cannot show how a treatment works. The waves, their order and the assumptions of the mediation model are checked before it is fitted.
Baseline Imbalance in Trials Baseline differences between randomized arms arise by chance, so they are not tested for significance. The baseline score enters the model as a covariate, which gives a more precise comparison than change scores.
STATISTICAL METHODS
Methods Selected for the Research Question and Study Design
Each method family follows the outcome and the data structure of your treatment study, and the line under each title explains why it suits clinical psychology data.
Longitudinal & Multilevel Analysis
Repeated symptom scores with missed visits are modeled with every available observation, and patients are nested within therapists, instead of testing each time point separately.
Comparative & Hypothesis Testing
The end-of-treatment difference between arms is estimated with the baseline score as a covariate, and an uncontrolled pre-post change is reported with its interval.
Multivariate, Scale & Reliability
Reliability is estimated in your sample, and for clinician-rated measures agreement between raters is quantified, so a change in the score can be read against measurement error.
Regression & Predictive Modeling
Remission and response are binary outcomes, modeled with a few prespecified predictors and penalized estimation when events are few, and a mediator only where its timing allows.
Bayesian & Computational Methods
Missing visits are handled by multiple imputation, and weakly informative priors can stabilize the estimates of a small trial, with the prior stated and its influence examined.
Clinical, Survival & Epidemiological
Time to relapse or to leaving treatment is analyzed as a time-to-event outcome, since patients still well at last contact are censored, not treated as non-events.
DELIVERABLES
Statistical Outputs Prepared for Your Research
The outputs follow the structure of a treatment study, from the review of the assessment schedule to the scripts that regenerate every reported value.
Data & Study Review Scoring manuals, the schedule of assessments, therapist and site identifiers and the pattern of missed visits are reviewed, so every score is traceable to a patient and a time point.
Statistical Analysis The agreed mixed models or covariate-adjusted comparisons, classification of reliable change, sensitivity analyses for dropout and diagnostics for the outcome model.
Effect Estimates & Uncertainty Between-group differences, standardized effects such as Hedges' g, the proportions with reliable change and reliability coefficients, each with a 95% confidence interval.
Tables & Figures Participant flow, baseline and outcome tables, symptom trajectories by group and the counts of reliable improvement and deterioration that treatment studies are expected to report.
Statistical Interpretation Findings are explained in clinical terms, with the limits of the design stated, such as a small sample, therapist clustering or missing follow-up, and the wording kept within what it supports.
Analysis Files SPSS syntax, R or Mplus scripts, with the scoring rules and the coding of missed visits documented so that each table can be regenerated.
REPORTING
Reporting Clinical Psychology Research
The guideline depends on whether treatment was randomized, and each is applied only where it fits your study.
A randomized therapy trial reports to CONSORT 2025 and its protocol to SPIRIT 2025, while an evaluation without randomization follows TREND. Manuscripts for psychology journals also give the APA 7 and JARS-Quant account of the sample, exclusions, measures and analysis, and a study of an instrument’s measurement properties follows COSMIN. Participant flow is tabulated from the analysis, dropout is reported by arm with the way it was handled, and the primary outcome is reported as it was registered, with any change stated. Each symptom change is shown as a difference with its interval, an effect size and the proportion with reliable change, so the manuscript and the output agree.
ILLUSTRATIVE EXAMPLE
How the Analysis May Look in Practice
An illustration of how a treatment comparison with dropout moves from a protocol to a defensible, reportable result.
ILLUSTRATIVE EXAMPLE · OUTPATIENT ANXIETY TREATMENT RESEARCH · UNITED STATES
An outpatient clinic compared a ten-session cognitive behavioral program with treatment as usual in a randomized trial, with an anxiety scale completed at baseline, mid-treatment, the end of treatment and follow-up.
Patients who felt worse were more likely to miss later visits, and three therapists delivered both treatments. Before outcomes were examined, we fixed the analysis in writing: all randomized patients in a linear mixed model with baseline score as a covariate, a random intercept for patient, therapist entered as a fixed effect because three therapists are too few to estimate a therapist variance, and the end-of-treatment difference as the primary estimate. Reliable change was defined from the scale’s reliability in the trial sample, and a sensitivity analysis shifted the imputed values of patients who dropped out. A process measure taken only at the end of treatment was reported as exploratory, not as a mediator.
Outcome: The group received the adjusted between-group difference with its 95% confidence interval and Hedges' g, the proportion with reliable improvement in each arm, CONSORT 2025 flow counts, a methods paragraph and the scripts.
This example does not describe a real client or project. We never share details of actual client work, in line with the confidentiality we promise every researcher. This illustrative example is included to show the type, scope, and standard of work involved in this kind of analysis.
RESEARCHER VOICE
What Researchers Say
Feedback from a clinical psychology researcher on a small sample with unstable estimates, and on a team that already understood the design, the data problems and the reporting expectations.
“What worked for me was that I did not have to translate the whole Clinical Psychology context before discussing the statistics. They understood the study design, the usual data problems, and the reporting expectations. We were particularly concerned about a relatively small sample and unstable estimates, and that was dealt with properly from the start.”
HOW WE ASSESS YOUR PROJECT
We Review Your Actual Research
Your assessment is based on a review of your actual research materials and analytical requirements, not a generic project category.
You Submit Your Project Details
Share whatever is available at the current stage, such as the research aim, protocol or proposal, dataset, variable list, manuscript or statistical requirements.
We Review Your Research and Define the Scope
Your statistically trained Research Coordinator reviews the materials to understand the study design, research questions, data structure, analytical requirements and expected deliverables. If clarification is needed, additional documents or information can be requested before the assessment is finalized.
Your Analysis Is Completed and Delivered
Once the scope and fixed project fee are agreed, the in-house statistician completes the work and the Research Coordinator remains the communication point through delivery, with the analysis, deliverables and timeline as agreed.
SOFTWARE
Statistical Software for Clinical Psychology Research
Treatment studies are analyzed in the software your clinic or department already uses, with syntax or scripts provided where applicable.
Statistical Support Within Your Research OnlineSPSS provides the agreed methodological, analytical, and statistical reporting support. The researcher or research team retains the scientific question, substantive interpretation, conclusions, and overall responsibility for the research.
FREQUENTLY ASKED QUESTIONS
Questions About Statistical Analysis for Clinical Psychology
Short answers to what clinical researchers ask before sharing a study, from method choice and completers to planning, review and confidentiality.
What statistical methods are used in clinical psychology research?
Treatment studies most often use mixed-effects models for repeated symptom scores, covariate-adjusted group comparisons, reliable change classification and logistic models for response or remission, chosen by the design and the pattern of dropout. Our Statistical Methods pages show how each family is applied.
Should I analyze only the patients who completed treatment?
A completers-only analysis usually breaks the comparison that randomization created, because who completes is often related to the outcome. The analysis keeps every randomized patient and uses a model that draws on all available visits. Where the question is about patients who completed, it is labeled as such and the difference from the full analysis is shown.
How many patients does my treatment study need?
The sample follows the precision needed for the primary outcome, the correlation between repeated scores and the expected dropout. For a protocol we calculate it for that outcome and write it into the analysis plan. For a small pilot we describe what precision the feasible sample gives, and the study is reported as a pilot, not as a test of effectiveness.
Can you review a mediation analysis I have already run?
Yes. We review the order in which the mediator and outcome were measured, the measurement of each, the model and its assumptions, and the sample size against the number of paths. The review states which conclusions the data support, and an indirect effect from a cross-sectional model is worded as an association.
How is confidential clinical research data handled?
Study datasets, protocols and analysis files are handled confidentially, and project information is not published. An NDA is available where required and can be arranged before sensitive materials are shared. Read more about confidentiality and data security.
What happens when I request an assessment, and how long does it take?
You share your research question, design and relevant materials, and a statistically trained Research Coordinator reviews the request and responds the same day. The analysis, deliverables, timeline and one fixed project fee are then confirmed before any work begins, so the timeline follows from the agreed scope.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
Discuss Your
Clinical Psychology Research With Us
Share your research question, design and available data, even if only a protocol or a scoring manual exists. We review how dropout and change will be handled and confirm the approach before any work begins.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014