Hi everyone! Today's lecture video will discuss the scientific methods in psychology. Three topics pertaining to scientific methods in psychological research discussed in the video include scientific thinking in psychological science, research process in psychological science, as well as evaluations and indicators of scientific theory in psychology.
Psychology as a science seeks to explain various psychological phenomena, namely thoughts, emotions, and behaviors. Psychological science applies systematic methods to observe human behavior and to make inferences based on these observations. The accumulation of these inferences would then be used to generate a psychological theory that aims to explain certain psychological phenomena. As part of science, The explanations proposed by a scientist in psychology is an explanation that is, of course, based on the principles of science. The word science itself comes from the Latin word that means knowledge.
Science and scientific theories are different from layman's theories because science, by definition, is a search for knowledge based on carefully observed and replicable data. In this diagram, you can see the stages of research. In general, scientific research begins with the formulation of a hypothesis based on observations or general assumptions or statements, such as the statement that children tend to imitate the behaviors they see.
For example, when researchers are interested in the impact of video games on the behavior of school children, they will first observe children playing video games, and documents or record specific behaviors displayed by the children, including the exact forms of the behavior, how often the behavior is displayed, and when the behavior is displayed. After perhaps days or weeks of observations, researchers may suspect that watching violent video games will lead to aggressive behavior. Good observations and measurements often show patterns that lead us to good hypotheses as well. And what is meant by a good hypothesis here is simply a clear, predictive statement.
For example, the researchers could predict that watching violent scenes in video games will increase children's aggressive behavior, or reduction in the amount of violence in video games would decrease aggressive behavior in children. Next, the hypothesis needs to be tested. Hypothesis testing requires systematic methods.
So what is the proper method for testing hypotheses? We need to first imagine every method that's available to us. If the hypothesis predicts watching violent scenes in video games will increase children's aggressive behavior, then one way to test this hypothesis is to check whether children who play video games full of violent scenes also behave aggressively towards others. But this method may not really help our inferences and explanations.
Since we can't be sure whether the cause of this aggressive behavior is indeed violence in video games, or whether it's caused by other things, therefore, we need to implement other methods that would help us make better inferences. For example, we can recruit a group of school-age children to participate in a video game competition. Then, we divide the children into two groups. The first group can play violent video games.
while the second group will play video games that contain no violence at all. Using this method, we are essentially manipulating the presentation of violence in the video games. Following the video game competition, we can observe and compare the frequency of aggressive behaviors in the children in the two groups. If our hypothesis is correct, then we should see more aggressive behavior in the first group than in the second group. The strength of this method is that we can reasonably conclude that it is in fact violence seen in video games that affects children's aggressive behavior.
But there are also disadvantages to this method. In this method, we can only expose the children to violent video games for one day or a few days at best. So our inferences only apply to conditions where children play video games for a few days only.
Yet, in real life, children don't only play video games for a day or two. In other words, we cannot be sure. whether a consistent increase in aggressive behavior can be expected from longer-term exposure to violent video games. Each method obviously has its own strengths and weaknesses. However, if a phenomenon is investigated using various methods and research in the area consistently produces the same results and the same interpretations, then the researcher's confidence in the hypothesis will also strengthen.
The fundamental issue in many psychological studies pertains to behavioral measurement. Measurements of behaviors need to be carried out with clear definitions and rules, especially considering that a lot of psychological phenomena, such as violent behavior, are extremely difficult to measure. Having clear definitions and rules of measurement will lead to more conclusive results and less ambiguous interpretations of the results of a study, including whether the results are found simply due to chance, or whether they indeed reflect the acquisition of novel explanations about the phenomenon that is being studied.
A researcher's final objective is to interpret the results of the research. If the obtained results contradict the prediction made in the hypothesis, then the researcher must abandon or modify this hypothesis. However, if the results match the hypothesis, then this would, of course, increase the researcher's confidence in his or her prediction. Nevertheless, a good researcher would still need to consider other hypotheses that could lead to the same finding. Replicability is an important issue in psychological science.
Research results that corroborate a researcher's belief in the hypothesis are said to be replicable, which means that the same results would also be obtained by others who use the same methods, same definitions, and same procedures of research. To ensure replicability, anyone who publishes scientific research is expected to report in great detail the methods and procedures used in their study, to allow others to replicate the study. Anyone who misrepresents their research methods risks losing the trust of fellow researchers as well as the public.
Many research results are obtained through studies that employ a small number of participants, and this will naturally weaken the interpretation of the results or reduce the statistical effect obtained. However, as long as the reported research methods are clear, and the results are consistent with what exists in the literature, the science community will generally allow for such results to be published. To ascertain confidence in the results of studies that find small statistical effects, scientists often resort to the use of meta-analysis. Meta-analysis is a method that analyzes the results of many studies all at once, treating them as a way of understanding the results of the study.
as if they came from one large study. Meta-analysis can uncover the extent to which study procedures can provide replicable results. If the meta-analysis determines that these small effects are not replicable in other studies, then the results of the study would usually be deemed obsolete. This rule may sound a bit harsh, but it is nevertheless the best way to prevent errors in scientific reasoning.
Now let's look at the status of psychology as a science. Is it a science or is it pseudoscience? You've learned that data that can be replicated to support multiple hypotheses eventually increases researchers'confidence in these hypotheses.
This is because the hypotheses help us to deduce and to explain psychological phenomena. Explanations built from replicable data will eventually be used to formulate a theory. Remember that a theory is no guessing game. A scientific theory is an explanation or a model that corresponds to various observations and produces accurate predictions. A good theory starts with as few assumptions as possible and leads to many correct predictions.
The use of as few assumptions as possible is important. Because what the general public needs is a theory that is simple. Simple enough, yet applicable in various contexts. So how do we evaluate the quality of a theory?
Four indicators are generally used to evaluate whether a theory is a good one, or whether it's simply pseudoscience. The first indicator is replicability, which we've already discussed previously. A theory is considered scientific if it is supported by replicable results. A theory is also scientific if it is falsifiable or can be demonstrated to be false.
A good theory needs to be formulated in clear enough statements to allow scientists to seek for or to imagine real-life scenarios that would otherwise prove the theory to be false. If there is no way of imagining a piece of evidence or an observation that can falsify the theory, then that theory does not fulfill the requirement of a good theory because it would render the theory too abstract, lacking in clear predictions, and therefore likely to be inaccurate. The next indicator of a good theory is that it needs to fulfill the burden of proof. Burden of proof refers to the obligation of supporters of a theoretical claim to present evidence that would support the claim. That is to say, that so long as we have faith in a claim, then we are obliged to demonstrate some empirical evidence that would strengthen the claim.
The last indicator is the principle of parsimony. In science, a theory that consists of as few assumptions as possible, especially those that have been used by established theories, will be judged to be simpler and more consistent with others. other pre-existing theories. This principle of parsimony is also known as Occam's razor.
The principle of parsimony is rather conservative because it encourages us to rely on assumptions and theories that clearly work and to avoid as much as possible new assumptions that have not yet been clearly substantiated.