Department of Biostatistics
Training and Workshops
The Department of Biostatistics at PILL has been regularly conducting workshops on various topics including Fundamentals of Biostatistics, Basic and Advance Level of Statistical Package (SPSS) and AMOS; including, Correlation, Regression Analysis, t-test, ANOVA, ANCOVA, MANOVA, Factor Analysis, Mediation and Moderation Analysis.
Department Research
Our researchers conduct methodological research across a broad range of areas, including design and analysis of clinical trials, survival analysis, longitudinal analysis, semiparametric methods, missing data, causal inference, measurement error, surveillance methods.
What We Do:
- Prepare research findings and presentations
- Research data archiving
- Grant preparation / study design
- Determine statistical methods
- Sample size determination
- Electronic data capture
- High-dimensional data storage
- Multicenter study coordination
- Data and safety monitoring
- Data analysis and cleaning
- Manuscript review
Psychometric Properties
Psychometric properties are important characteristics of any psychological assessment tool or instrument because they provide critical information regarding its reliability and validity. Reliability relates to its consistency and stability, exhibiting its ability to produce identical results when conducted again under the same settings. Validity, on the other hand, evaluates whether the instrument measures the desired construct and its ability to predict or explain relevant outcomes.
Other psychometric properties, in addition to reliability and validity, include norming. Norming is creating a standardized reference group for comparison and responsiveness, which measures the instrument’s sensitivity to detecting changes over time. Understanding and assessing these features is critical to assure the validity and use of psychological evaluations in a variety of areas, including education, clinical psychology, and human resource management.
Types of Reliability
- Test-Retest Reliability: This measures the consistency of scores when the same test is given to the same set of people on two separate dates. A high test-retest reliability shows that the results are consistent across time.
- Internal Consistency Reliability: This assesses how well the items in a single test or questionnaire are connected to one another. Cronbach’s alpha and split-half reliability are two common indicators of internal consistency.
- Inter-Rater Reliability: The consistency of observations or judgments made by different raters or judges is assessed in this sort of reliability. It is frequently employed in situations involving subjective evaluations, such as behavioral assessments.
- Composite Reliability is a statistical measure used to assess the internal consistency or reliability of a measurement scale or instrument that consists of multiple items or questions designed to measure the same underlying construct or dimension. It is particularly relevant in the context of psychometrics and survey research.
Types of Validity
- Content Validity: This form of validity examines whether a measurement instrument sufficiently represents the complete domain of the construct under consideration. Experts frequently assess information to ensure that it is thorough and useful.
- Criterion Validity: Criterion validity measures a measurement instrument’s ability to predict or correlate with a gold standard external criterion. There are two kinds of subtypes:
- Concurrent Validity: This assesses how well the instrument correlates with a criterion measured concurrently.
- Predictive Validity: It assesses the instrument’s ability to predict future performance or behavior.
- Construct validity determines whether a measurement instrument accurately measures the underlying theoretical construct that it claims to measure. It frequently entails investigating the relationships between the instrument and other constructs in order to support its validity.
- Convergent Validity: The degree to which an instrument correlates with other instruments or measures that are theoretically predicted to be related to the same construct is measured by this form of validity. A high convergent validity value suggests that the instrument is measuring what it is designed to measure.
- Divergent validity (Discriminant Validity) examines an instrument’s associations with unrelated constructs to determine whether it is distinct from other constructs. It ensures that the equipment is not measuring something that it should not be measuring.
Biostatistics News:
Under the supervision of Dr Steven Lane, the Department of Biostatistics is also going to introduce the three months’ internship and certificate program from April 2024 in the field of Biostatistics for undergraduate, postgraduate students and health professionals (doctors, nurses, psychologists, psychiatrists and other mental health professionals).
