According to Wikipedia, Internal validity is a property of scientific research that demonstrates the degree to which a causal conclusion based on a research is warranted. Such warrant is constituted by the level to which an investigation minimizes systematic error (or ‘bias’). During research design, the threats to internal validity consist of inadequate knowledge of, or contradictions in the logic.
What are the Threats to Internal Validity in Research ?
Instruments don’t measure what they purport to measure: It means the findings are invalid. For example: The measurement approach or instrument is not tested for dependability; that is, distinct experts or evaluators observing identical things would NOT get the same scores.
History: Events outside the experiment or between repeated measures of the dependent variable might influence participants’ replies to experimental procedures. Frequently, these are major events (natural tragedy, governmental change, and so on.) which influence participants’ perceptions and behaviors in a way that it becomes difficult to find out whether any change on the dependent measures is due to the independent variable, or the historical event.
Maturation: This describes changes which normally take place with time (e.g., power, increasing knowledge). For example, let’s say a new approach to improve children’s attention span is examined in an experimental intervention. And let’s say that most kids are more attentive 8 weeks later, during the post-test. The experimenter may believe this progress is the response to the method, but the rival hypothesis is simply that the kids turned more mature, and this increased their attention. To handle the extraneous variable of maturation, experts make use of equivalent comparison groups, or utilize experimental designs where the experimental group serves as its own control (e.g., the equivalent time samples design).
Testing: Repeatedly measuring the individuals can result in bias. Participants can keep in mind the right answers or may be conditioned to know that they are being tested. Frequently taking (the same or similar) intelligence tests generally results in score gains, but rather than concluding that the underlying skills have changed for good, this threat to Internal Validity offers good rival hypotheses.
Regression: An individual’s performance of any activity deviates within a particular range. In research, a group’s pre-test efficiency might (by chance) be abnormally low or high; quite a few people have a good day or a bad day. On later testing, the group’s performance regresses to the mean (i.e., is more usual). The investigator may erroneously treat variations between pre- and post-test results as the result of an intervention or as the failure of an intervention, when in fact, the group just turned in its average or usual performance.
Mortality/differential attrition: This means the differential loss of individuals from comparison groups. For instance, an experimental intervention may seem to work simply because individuals with whom it was not going to work dropped out. In the same way, an intervention may seem to work no better than nothing at all, simply because individuals in the group who would have gotten worse over time dropped out, leaving individuals the control group who improved. Therefore, the group scores about the same as the experimental group.
Diffusion: If treatment effects pass on from treatment groups to control groups, a lack of differences between experimental and control groups could be noticed. It doesn’t mean, however, that the independent variable has no effect or that there is no connection between dependent and independent variable.
Compensatory rivalry/resentful demoralization: Knowing they’re in a control group, a few members try to change by themselves. Improvement in the control group could be mistaken to imply that the intervention is no better than no intervention.
Threats to internal validity in research can occur through the entire research process. A good research design is definitely of crucial value when seeking high internal validity.
Speak Your Mind