Research-related questions

How do I define my study and research question?
CSTAR is available to assist with defining and planning a study.  This includes constructing primary and secondary research questions, identifying experimental designs which address the proposed questions, developing a data dictionary for the variables to be studied and a statistical plan outlining how the data collected will be analysed.

What is the difference between primary and secondary research questions?
Particularly in experiments with an intervention, there is generally one primary question the study intends to answer.  As part of the study, ancillary data is often collected and studied to answer secondary questions.  These questions generally are instrumental in defining future research projects.  The study design is centered on the primary question and any causal claims of the study are specifically in regards to the primary question.

What type of study meets my objectives?
There are many classes of studies which meet a variety of objectives.  CSTAR has analysts trained in a broad variety of quantitative and qualitative research methods and can assist with selecting an approach appropriate to your objective.  This includes identifying if an observational study is sufficient or if an experiment with an intervention is necessary.  These types of decisions can be made based on whether the objective is to make observations about a specific case, describe a group of individuals, or generalize results to a population.

Can CSTAR help me with a sample size/power calculation?
CSTAR can help researchers estimate the power of a study being proposed.  Power calculations estimate how sensitive an experiment is expected to be based on an effect size and the proposed sample sizes.  As such, it is based on what a researcher expects to happen during a study.  Optimally, the researcher has an estimate of the effect size anticipated in the study.  The effect size is a function of the expected change in response and the variability of the measures of interest.  Often, if the ranges of scores for the various treatment groups are available, estimates can be made leading to an estimate of the power expected for a study.

How can the effect size of a treatment be estimated when there is very little information?
Power calculations have three components; power, effect size and sample size.  As long as two of the three components are known, the third can be determined.  Therefore, if the sample size and power are fixed, a minimum detectable effect size can be determined.  This can often give a sense of whether or not a study will have sufficient power to be worthwhile.  Similarly, if the desired effect size and power are known, the required sample sizes can be established.

Which experimental design is right for me?
CSTAR is able to assist in selecting an experimental design to maximize the efficiency of the study.  Based on the resources available, the research questions and the weights of the questions, advanced experimental designs can be explored to reach an optimal design while reducing costs especially for complicated studies. 

How and what do I measure?
These are issues that are often overlooked but they are of great importance in study design.  A data dictionary is an excellent resource which identifies what traits are being measured, how the trait is being measured and the nature of these measurements.  CSTAR is well versed in ways to observe and analyse both physical measurements (e.g. height, weight, blood pressure, serum levels) as well as conceptual measures (e.g. quality of life and depression index). 

Can CSTAR help with my qualitative research?
CSTAR has experience in conducting and analysing qualitative studies.  Many research issues, such as sampling and randomisation, are applicable to all types of studies, and CSTAR can help with the study design to maximize a study’s validity.  Additionally, CSTAR can assist with operationalising a study and developing methods of quantifying subjective data.  The Centre can also assist in reporting the reliability of any instruments such as surveys or between raters in multi-rater studies.

How do I design a questionnaire?
If a standard scale for a trait is not available, researchers can develop questionnaires to access or measure a non-physical trait.  The Centre has experienced researchers who can assist and advise researchers in developing a questionnaire and evaluating how effective the questionnaire is at measuring the desired trait.

Is my questionnaire reliable?
Whether using a standard scale or a developed questionnaire, certain properties of the questionnaire, such as its reliability should be reported.  CSTAR is in a position to advise on which measures should be reported and how to calculate these measures to report that the scale is performing as intended.

Is my questionnaire valid?
When developing a questionnaire, it is important to establish a sense of validity in the instrument.  CSTAR can advise on how to use secondary outcomes, multivariate techniques and qualitative research methods to build an argument as to the validity of a developed questionnaire.

I’ve collected this data, what does it tell me?
Analysts are available to assist with previously collected data - whether the required assistance is discerning what story the data is telling or identifying which research questions can be answered.