If a sample for an experimental research is collected based on non probability sampling( convenience and probability). Now the sample has to be divided into the groups intervention and control group
It should be done on the principles of True experimental design or Quasi-experimental design?
Lastly how do we assign the participants non-randomly in quasi-experimental designs as all the participants share the same characteristics.
True experimental designs are preferable to quasi-experimental designs, so if you can randomly assign participants to an intervention versus a control group, that would be your best option.
David L Morgan
Does this mean that if we have used non-probability sampling in our research, we must use the quasi-experimental design because we have not randomly selected the participants?
Secondly, how can we assign participants non-randomly in quasi-experiments?
There is a key difference between random sampling from a population versus random assignment to an experimental group or a control group. It is the second of these that determines whether you are conducting a true experiment.
With regard to your second question, quasi-experiments are only done when random assignment is not possible. In that case, assignment is usually based on some set of pre-existing sources (e.g. schools or hospitals).
David L Morgan
Thank you, David!
Finally, can you please advise me on which experimental design would be more appropriate for my research? I am providing interventions to my research participants to reduce the effects of betrayal. The sample has been accessed through academic institutions based on non-probability sampling. I think both experimental designs could be administered but I am unable to take into account the complexities of choosing one and appropriate
design.
To address your first question, it's important to consider that true experimental design involves random assignment of participants to either the intervention or control condition. You may have a sample obtained through convenience, for example, a group of university adults voluntarily participating in Study X; what determines if it's experimental is whether each of these voluntary participants has an equal chance of being assigned to either the experimental or control group9
Quasi-experimental designs involve the allocation of participants into intervention and control groups without random assignment. Instead, participants may be assigned based on existing characteristics or criteria, such as demographics or geographical location. While quasi-experimental designs lack the randomization seen in true experimental designs, they are often employed when random assignment is impractical or unethical.
Momina, assuming you have 100 participants with similar demographics and you just decide that the 1st 50 people will be one group and the 2nd 50 in another group, that is non-random assignment, which is different from using probability framework to randomly assing them.
Erika Margarita Arias
Thanks a lot, Erika for the detailed answer.
What would you recommend would be an appropriate approach for dividing the sample equally and randomly for true experiment design?
- Badges
- Science topic
- Similar topics
- Social Science