SQ9

Please describe the difference between experimental design, pre-experimental design, and quasi experimental design.

Explain why you might use one design type over another (you don’t have to describe this for all three but consider what might be a benefit of one design type in a certain situation). Also consider the possible threats to internal validity that can occur with each design type that may impact your choice. (In other words, it might be best to come up with an example and describe which research design you would use and why.)

27 thoughts on “SQ9”

  1. In an experimental design, there is A pre-experimental design there are multiple explanations for the obtained results due to the method of data collection. In a quasi-experiment design, an independent variable is also manipulated but the groups are not equivalent. I believe that in some cases, a quasi-experimental design may be the best choice. Due to the fact that a quasi-experiment makes efforts to account for an limit possible threats to internal validity, I believe that researchers can obtain the most reliable data. I find that in many circumstances, confounds such as the history effect, the maturation effect, or the testing effect can be best accounted for if one chose to use a quasi-experimental design.

  2. An experimental design is where two (or more) equivalent groups of subjects are exposed to different independent variables. The differences in the dependent variable can be attributed to the differences in the independent variable. The benefit of using this design is that causal inferences can be made when using this design. In a pre-experimental design, multiple alternative explanations are present after collecting the data from one group that is tested twice (pre-test/post-test). For example, outside factors may be an explanation for the results found. Quasi-experimental designs are used when there are non-equivalent groups present and an independent variable is manipulated. This design could be used to compare differences in subject variables (for example, gender and age).

    1. Hi Simona,
      I agree that quasi-experimental designs can be a powerful tool when groups are not equivalent as compared to experimental designs. Quasi-experimental designs are usually conducted in natural environments and therefore lack the artificiality that comes with controlled experiments in a laboratory. I see this as another advantage that can facilitate generalizations about the population.

      1. Yes Jessica, I agree that an advantage of quasi-experimental designs are that they are usually conducted in natural settings which are good for studies that are looking to observe behavior rather than test the manipulation of behavior.

  3. In an experimental design, groups of participants are randomly assigned to different conditions (or variables) in a research experiment. A pre-experimental design also follows the steps of an experimental design however it does not include a control group. This means that the researchers do not compare the results between treatment and non-treatment groups, making it difficult to draw conclusions from the data. A quasi-experimental design is also similar to an experimental design but the major difference is that it does not randomly assign participants. The purpose of this design is to identify causal impacts of an intervention. There are multiple threats to internal validity that can be drawn from the use of any of these designs. For example, a common threat to internal validity is subject attrition. This is when participants choose to withdraw from a study or when their data are chosen to not be used, making the groups unequal. High drop-out rates effect the validity of the study. Because quasi-experimental designs already incorporate inequivalent groups, studies using this design are more likely to experience threats to internal validity. From this perspective, it might be more convenient to use an experimental design.

  4. In an experimental design, participants are randomly selected and randomly assigned to two groups, the experimental group and the control group. The control group is where the independent variable is not manipulated, while the experimental group is where the independent variable is manipulated. The pre- experimental design is similar to experimental, however it lacks a control group and does not have random sampling. Quasi experimental on the other hand, is similar to experimental design but lacks random assignment. I believe using experimental design may be the best choice because it manipulates one condition, while controls the other. Having two groups, can help researchers measure and compare both groups, making sure there are no confounds affecting results, and if there were, it will be easier to locate it. Possible threats to internal validity for quasi experimental is that there are no random assignment, so certain people are chosen for the experiment, this can cause bias, and affect the results, for that reason I would rather use an experimental design.

    1. Hi,

      Yes, agreed! Although I chose a quasi-experimental design as my choice of design, in an experimental design it will be easier to see which variable affects the other and we will be able to locate any confounds that are found since the groups are randomly assigned and equal.

  5. An experimental design involves using a group of randomly selected participants to partake in a certain task or under select conditions to measure the effect of a particular variable. A pre-experimental design is similar to a experimental design because it involves the collection of data but there is no control group to measure the data against which means that the results of the pre-experimental design can have various explanations and draw no real conclusions. Quasi-experimental designs are used when a true experiment cannot be carried out. They are similar to experimental designs as well except they do not have randomly assigned participants; the inability of the research to randomly assign conditions to participants is what causes the researcher to use a quasi-experiment instead of a real experiment. The comparison groups in a quasi-experiment are not equivalent. Some threats to internal validity in a quasi-experiment include subject attribution (since the groups are already uneven and losing data does more harm in a quasi-experiment than it does a real experiment) and selection bias within groups. To choose the correct design, one would have to consider the presence or absence of random assigning and whether or not there is control group to choose between these three designs.

  6. Hello,

    An experimental design refers to the type of experiment that exposes the subjects or participants to different independent variables. On the other hand, a pre-experimental design is when drawing any conclusions results difficult, if not impossible, due to a numerous of findings obtained from the data. As for a quasi-experimental design, is when the researchers manipulate the independent variables. However, the groups are not the same. I believe that, of these three study designs, the best one is the experimental design because by exposing the subjects or participants to different independent variables, the chances of gathering more significant data are higher. Also, by having two different groups, the chances of confounds are lower.

  7. A quasi-experimental design is the same as an experimental design, however, the difference is that there is no random assignment or equivalent groups. In a pre-experimental design, the data is collected in a way that there are alternative explanations for a study. This is due to data being collected from one group before and after a study, there is no control group involved. A pre-experimental design can have outside factors and alternative explanations rather than one explanation for the results of the study. In an experimental design, the participants are randomly assigned to a control and experimental group. For example, if I were to research the effect of alcohol on someone’s behavior before alcohol and after, I would use a quasi-experimental design because I will be able to compare once before and after the alcohol. However, a possible threat to the internal validity can be the assignment of my participants. Since they aren’t randomly assigned in a quasi-experimental design, the groups aren’t equal and we wouldn’t be able to know if the results are because of the alcohol or any extraneous variable. Another threat to the internal validity might be selection bias. A selection bias can occur if I only choose people that drink and people that drink regularly. This will skew my results and the reasoning behind my results will not be valid.

  8. An experimental design enables the researchers to test their hypothesis by reaching valid conclusions about relationships between independent and dependent variables. This is done by deliberately changing one or more variables in order to observe the effect the changes have on one or more response variables. (the IV on the DV). A benefit of using such is design is that we may determine causal inferences as opposed to others that may only look at the relationship between variables. Within an experimental design may be a pre-experimental design yields results for which there are several alternative explanations. Although, they follow the basic follow basic experimental steps but fail to include a control group. Lastly, a quasi-experimental design compares non-equivalent groups. For such design, there may be many threats to internal validity such as history effects, maturation effects, testing effects, subject mortality etc. As researchers, it is important to minimize these threats in order to ensure the best results.

  9. An experimental design requires two criteria to be met. First participants must be randomly assign to groups. Also,there should be at least one experimental and one control group. Quasi-experimental designs are used in place of true experiments when it is impossible to randomly assign participants to experimental conditions. An example being that a researcher can not randomly assign a person to be male or female if they have already been assigned a sex from birth. Threats to validity include maturation effects, selection bias and testing effects to name a few. In a pre-experimental design a single case is observed before and after treatment. This can also be done with groups where one group has treatment and the other group doesn’t have treatment and both are compared. In both a group and single case changes are assumed to be a result of the treatment making it impossible to rule out alternative explanations affecting the validity of results. Of the three designs, in most cases I would use the experimental design as it controls extraneous variable by having a control group present for comparison with manipulated variables.

  10. The key differences between experimental, pre-experimental, and quasi-experimental design are that an experimental design is the random assignment of participants/subjects to groups with a control group and an experimental group. A pre-experimental design lacks a control group and a quasi-experimental design lacks random assignment. I think it would be best to use a quasi-experimental design in most cases because it allows for the experimenter to delve into the topic he/she wants to study due to the fact that random assignment is eliminated. A quasi-experimental design is also more practical as compared to the experimental and pre experimental designs. It takes less time to find participants and threats to the validity of the experiment can be found and corrected. The only significant limitation would be in the generalizability of the experiment. Since random assignment is eliminated in a quasi-experimental design, it is often hard to generalize the results of the experiment.

  11. There are three types of experimental designs: experimental, pre-experimental and quasi. An experimental design consist of random participants constructed into randomly assigned groups in order to test the effects of a particular variable. Whereas, pre-experimental design although is similar to an experimental design, it does not consist of a control group for comparison purposes. Additionally, the data collected is thus for analysis and not necessarily comparison due to the control group is absent. The quasi-experimental design does not include random assignment, but does include a control group where a variable is manipulated in order to gain results of the effects. The quasi-experimental design includes further designs such as: nonequivalent-control-group design, time-series design and multiple time-series design. Within a quasi-experimental design, there are many threats to internal validity, one of which is selection bias. Selection bias occurs because of unequal groups, thus when attempting to attain data, it is difficult to retrieve more data without a certain criterion in mind. All in all, the experimental design is the most useful design to gain accurate, non-bias information. There are equal groups which are randomly assigned and the information is already outlined to foresee comparisons between data because of how the experiment is already set up.

  12. A pre-experimental design is a basic experiment but one thing it does not have is a control group. Without a control group there is nothing to compare the data to. An experimental design is where the subjects or participants are randomly assigned to a control or treatment group while the quasi-experiment design would be having the subjects not randomly assigned which would take away from the fact that it would not be a true experiment. The best option is the experimental design where there is both a control group as well as random organization of the participants so the data will come out to be valid.

  13. In a pre-experimental design, the opportunity for alternate explanations are a possibility. The reason for these alternate ideas comes from the fact that pre-experimental designs do not include control groups for the most part. They consist of a single groups being tested before and after the experiment is performed. However, in the regular experimental design, the process consists of assigns random people to both a control group and a test group. The process must also be repeated by other researchers. The benefit of this design is that it helps find accurate statistics for the study they are conducting. In a Quasi-Experiment the study lacks randomization and the control group and treatment group may have different terms. Quasi-experiments are able to provide several hypotheses.

  14. In a normal experiment, there are two groups, an experimental group and a control group. The independent variable is manipulated and the differences that occur are recorded. However in a preexperimental design there is only one group and that same is recorded before and after being exposed to an independent variable. Because there is no group control group an experimenter cannot conclude that the independent variable was the cause of the changes before and after treatment. In a Quasi experimental design the control and experimental group are not equal. However the control group receives the independent variable no different than in a regular experiment. It can be concluded that changes between the control and experimental treatment are the result of the independent variable however one must take into consideration that the group were not equal to begin with. Under normal circumstances a regular experimental design is best. However when there are ethical concerns, for example, in the case of determining the effect of prenatal care Quasi experiment can take to different set of women who are pregnant to compare

  15. An experimental design has its participants sorted into groups, usually a control and experiment group and are randomly assigned. A quasi-experiment is very similar to that of a regular experiment, except that its participants are not equivalent. A pre-experimental design does not have a control group.
    An issue with a normal experiment is that the participants can be too random to find any reliable result with them. However, this can also be said with quasi-experiments, but in this case, the results may be biased due to the lack of randomness. In a pre-experimental design, it can be difficult to find a group to compare results with, since there is only one group.
    An advantage of using one of the smaller designs over the complete but larger one could be due to logistical restraints. Simply lacking the number of eligible participants of a study can mean that a group of researchers are better off with having less people to experiment on.

  16. An experimental design contains 2 groups,experimental and control, where the experimental group is exposed to a manipulated variable while the control is not. The 2 groups are usually comparable because they are equivalent such that random sampling has occurred where any subject or participant has the possibility of being in either groups. Pre-experimental design does not have a control but rather uses a pretest and post-test methodology where observations are made before and after manipulation of a variable. Another aspect of pre-experimental design is the quantity of data that is collected, there is often a lot of it to allow for alternative explanations to conclusions regarding the independent variable. Because of this it is difficult to draw definitive conclusions from a pre-experimental design. A quasi-experimental design contains either 1 or more groups. When there are multiple groups the groups are not equivalent due to innate differences or due to sampling restriction. Designs within quasi-experimental may include pretest and post-test or multiple observations of the group(s). An experimental design would yield more direct results and conclusive conclusions. However, there are many ways in which groups of people differ due to innate characteristics (gender, ethnicity, even culture derived from living in a particular environment) and a quasi-experimental design is best because it both evaluates within groups and between. There are multiple threats to internal validity in a quasi-experimental because it lacks the lab controlled settings that experimental designs, maturation and testing effects are more present in quasi-experimental because of the length of time and the presence of a pretest and post-test design.

  17. In an experimental design, the participants are chosen randomly and assigned to random groups in order to test the effects of a particular variable. A pre-experimental design examines cause and effect, but has poor internal validity and also controls very few of the threats to internal and external validity. This type of design also does not have randomization. In a quasi-experimental design, researchers select the participants for different conditions from pre-existing groups but this type of design also lacks random assignment. I believe using experimental design may be the best choice because it manipulates one condition, while controls the other. This design also exposes the subjects or participants to different independent variables and chances of gathering more significant data can be higher.

  18. A regular experiment uses random selection and a control group to test a hypothesis. A pre-experiment is similar to a regular experiment except there is no control group. A quasi-experiement has a control group but does not use random selection. Each experiment design have their own uses. Quasi-experiment is useful when studying a specific group of people and for convenience. Using a pre-experiment is good for something simple such as comparing a students test result before and after conducting a certain teaching strategy. It saves times and may be used by everyday people to test something. A regular experiment will be ideal for the best results.

  19. An experimental design is one in which two or more equal groups are exposed to an independent variable. A quasi experimental design is basically an experimental design without random assignment. An independent variable is manipulated in this type of design, but equivalent groups are not possible. So, when a researcher is unable to have equivalent groups, such as looking at the differences between males and females, they resort to using a quasi-experimental design. A pre-experimental design is a type of quasi-experiment in which data is collected in a way that results in multiple explanations. This makes is very difficult to draw an accurate conclusion due to the fact that there are many outside variables that may have contributed to the results. I believe that if possible, the experimental design would be the best option when choosing a design. It allows researchers to draw causal conclusions.

Leave a Reply