Randomized controlled trial (RCT) is a plan experiment design used for analyze of efficiency of an intervention in human beings by comparing intervening group with control group. It is an experimental research type conducted for investigation using quantitative techniques for comparing outcomes of two groups. Generally, cluster samples of the same population is used.
In RCT, the investigation groups are the samples of those people who receive benefit from impose program, project, treatment or method. These groups are chosen from random using simple random sampling method. Another group is a sample who does not receive treatment and also taken from the same population. This group is called the control group which is also selected randomly from the same population. The data is collected during the different periods of time and the result is compared between the groups.
RCT is an experimental form of impact evaluation investigation where experimental or non experimental form of impact evaluations are analysis. In our city there may be more than one treatment groups depending on their it off methods used for target population.
For example, if there are two variety of treatment treatment A and treatment B then and there is atmos treatment methods. They are use of only treatment A, use of only treatment B and use of both treatment A and B. Therefore, if only one sample is taken from each treatment method then three sample group can be drawn. Sample group is considered as intervention group. And one extra sample is taken consisting of those people who doesn’t receive any kind of treatment from the same population. Therefore, in RCT, one sample is control group and maybe more than one group which is intervention group.
Let’s consider the example, I want to test the effectiveness of my teaching practice. I want to analyse the effectiveness of PowerPoint and case study method to a different sections of BSW class where I teach research methodology. My analysis is to find out the effectiveness of PowerPoint presentation, case study or both over traditional teaching lecture notes based practice. During the academic year I use traditional teaching base lecture for the students of section A of the morning shift, PowerPoint presentation method for morning shift section B students, case study method for day shift section B students and PowerPoint as well as case study method for Day shift section C students. During my session I conduct theoretical and practical test on a monthly basis. The result of all four sections are compute and compare. It is found that the average theoretical and practical marks of student of day shift section c is greater than other groups. So I can prove that teaching with PowerPoint as well as case study method is more effective method to teach the students. You may feel doubts that there is chances of choosing the bias samples but in knk international college the sections of a student are not assigned according to their rank but determine according to the time of admission and the average marks of a student of all the sections are almost equal.
since the random control trial is a quantitative methodology the sample size must be large enough to detect an effect of the treatment with sufficient precision.
Following are the main features of RCT:
- There are at least two groups that is control and intervening group. Control group does not receive any treatment but in triveni group receive treatment. Both groups are taken from the same population.
- It is a qualitative research design so use large sample size and statistical calculation for the measurement of progress.
- Initially, the outcome of both control and intervening groups are the same. What result may be different after the implementation of treatments.
- The measure outcomes of both control and intervening groups are changed during the period of treatment. This means that both are not static in nature.
Following are the main steps for conducting randomised control trial:
- Specify intervention (program,policy, project, method and their outcomes)
- Identify the target population and unit of assignment
- Randomly assign a sample of eligible population true treatment and control groups
- Collect the baseline data from both groups and develop a baseline report
- Collect midline data after the implementation of treatment and prepare report.
- Collect online data from both groups and calculate the impact
- The final report by comparing outcomes of both groups and draw the conclusion and recommendations are developed.
Type of bias in RCTs
In random control trial there may be many practical problems faced by the investigator. It leads to bias in RCT for comparison of outcome between groups. Some of these biases are discussed below:
- Selection bias
It occurs when decisions regarding the selection of the sample units are influenced by the knowledge investigator is occured. If the total sample unit is selected by chance and sub group it randomly than selection bias can be minimized.
- Allocation bias
There are several factors that affect the outcomes of intervention. If the groups are not randomly selected there is the chances of allocation bias. Selecting the different groups by chance is one of the methods to reduced allocation wise and balanced outcomes.
- Assessment bias
During intervention observations are made on the variety of outcome variables and does it at the end of intervention. Some observations are extremely objective and others are less objectives. If investigators know that the measurement of intervention of outcome variables contain some elements of subjective then there is the possibility of observations measurement is affected by knowledge of investigators. There is the possibility of assessment bias in this case.
- Publication bias
Ultimate objective of random control trial investigations is to develop a report on the basis of findings of observation. The submitted report are judged by export for publication of the article. The advice of exports and the number of other factors may influence the findings and conclusions.
- Stopping Rules
Once intervention is start the outcomes of it are affected by stopping rules. The duration of time and the target audience for the intervention is an important factor to reduce the bias. The sufficient allocation of time and target elements of population must be set for intervention.