Methods and issues in applied psychological research

¶ … Psychological Research

"It is difficult to turn the pages of a newspaper without coming across a story that makes an important claim about human nature" (America Psychological Association, 2003, par. 1).

Often, we come across specific claims about individual behavior, nature, principles, and/or dynamics which we might find interesting. These articles often cite research studies conducted on the subject matter, and more often than not, these articles also employ technical terms widely-used in research.

As students of psychology and the social sciences, we acknowledge the importance of research in this domain. It is through research that we get to learn more about our subject matter. It is through research that we obtain information about our topic of interest, which is human behavior.

Given the importance of research, not just in academic activity but also in our everyday encounters, this article finds it fitting to discuss some of the methods and issues surrounding applied psychological research.

Our discussion will open with the two types of statistics, i.e. descriptive and inferential statistics and their utility in applied psychological research. We shall then move to the issues surrounding sampling and research designs (true experiments/experimental designs and quasi-experimental designs).

Statistics: Basic Types and Research Considerations

Knowledge about the domain of psychology can be obtained and is developed by using scientific research methods. Such that psychology depends upon the observation of certain phenomena that occur in the natural world, it becomes dependent upon data that needs to be verified -- here comes the utility of statistics (University of York, n.d.).

Statistics is the defined as the "many methods for gathering sample data from a population. The main idea behind these methods is to be able to make some conclusion about the population based on" (Central Virginia Government's School, n.d., par.1).

There are two main types of Statistics, i.e. Inferential and Descriptive Statistics. In Descriptive Statistics, we normally attempt to measure something in which we assign numerical values based on well-defined rules. These rules give rise to data at four specific levels, i.e. categorical, ordinal, interval or ratio. The basic or preliminary step in research is to first organize the data according to the research design. Hence, psychologists use descriptive statistics to transform and obtain succinct data in either tabular or graphical form (Chow, 2002).

Inferential statistics, on the other hand, "…allows the research to make decisions or inferences by interpreting data patterns. Researchers use inferential statistics to determine whether an expected pattern designated by the theory and hypotheses are actually found in the observations" (Nachmias & Nachmias, 1996, p. 355).

From these conceptual definitions, we understand that descriptive and inferential statistics are statistical tools which help us in understanding our raw data. They both try to describe the relationships of our observations, however, inferential statistics goes as far as making "inferences" or analyzing future data patterns or behavior based on the current observed statistics (usually descriptive in nature) available.

Again, based on our conceptual definitions, one can see that descriptive statistics can be used to obtain organized or summarized information. But if one wants to know the likelihood of a particular behavior to happen, or one wants to know if a particular hypothesis can actually be observed in a particular case, then inferential statistics should be employed.

Sampling and Population

After finalizing our target respondents and respondent definition, we then move to our sampling consideration - this is when the understanding of population and sample plays a key role.

Population is defined as "a complete set of elements (persons or objects) that possess some common characteristics defined by the sampling criteria established by the researcher" (University of St. Louis Missouri, n.d., par.1). A sample, on the other hand, is known to be the subset of a particular population (Nachmias & Nachmias, 1996).

Population is represented by N. while sample is represented n (ibid). When do we use the population? Population is usually used when conducting census, when we need information from the universe of our defined criteria. However, it is understandably impractical to cover all the cases of your respondent definition as it is becomes extremely expensive. Aside from this, it is often impossible to cover each and every case as finding all the respondents will be very difficult (ibid). It is in this light that the utility of sampling technique is highlighted. Through various sampling techniques, one can get hold of respondents that can generalize our population, or more crudely, we are able to draw population estimates from our sample statistics (ibid).

On Research Design

A research design is known as the study's "blueprint" or the program that guides the investigator throughout the entire research process. It is a logical model of proof by which the researcher is allowed to make inferences from causal relations among the variables being tested (ibid). For the purpose of this section, we will be focusing on True Experiments or Experimental Design and Quasi-Experimental Design, particularly on their differences and validity issues.

Experimental vs. Quasi-Experimental Design

As we have already provided a discussion of what experimental design or true experiment is about, we will move on to the discussion of quasi-experimental design which we shall later compare to the former.

Quasi-experimental research design, as the name suggests, is as if or almost a true experiment. It usually involves the study of more than one sample, and often, over an extended period of time (ibid). In this kind of experiment, the word "trend" is utilized instead of "cause" or the causal relationship being established in true experiments (Austin Peay State University, 2006).

Its importance lies in the fact that it allows the researcher to conduct random selection of the samples from the population, but quasi-experimental studies do not require the researcher to randomly assign individual cases to the comparison groups. And even though this specific design has low internal validity as compared to true experiments, it enables the researcher to take on studies where the assignment of individuals to comparison groups might be unethical or impossible (Nachmias & Nachmias, 1996).

Summary & Conclusion

As we close this discussion, I believe that it is fitting to first provide a brief summary of what we have discussed so far, which will lead to the formulation of our conclusion.

We opened the discussion with the importance of statistics in applied psychological research, and with its two basic types, i.e. descriptive (provides a summarized and organized data) and inferential statistics (allows us to make decisions or inferences based on the observed behavior of the sample taken from the population).

As we have already mentioned sample and population, recall that our discussion highlighted the difference between the population (i.e. The entire set of relevant units of analysis or the universe of our respondent criteria) and sample (which is the subset of the population). Because it is often difficult, extremely expensive, and at times impossible to cover all the cases that are deemed to be relevant units of analysis, we instead use the sample.

We have also talked about the "blue print" of a study, or the logical model of proof which guides the researcher throughout the entire study -- i.e. The research design. It is by which the investigator determines the relationships between variables being tested. We have discussed true experiments, its nature and validity issues as well as quasi-experimental designs. We also provided a discussion of the difference between these two designs.

What is worth nothing, I believe, is the importance of a good research design. A good research design theoretically leads to good data. Hence, during the conceptualization of research design, careful thought needs to be employed.

Lastly, our research considerations, on which kind of statistical tool to use, whether to use sample or population, and/or which research design to employ, should always be anchored in our research objectives, on the things that we aim to know through the study.