Research Design and Research Methods - Assignment Example

RESEARCH DESIGN INTRODUCTION: This paper mainly talks about the basics of the research design and research methods and its types. The early part of the write-up includes the description and the differences between correlational and experimental research methods, while the latter part emphasizes on validity, reliability and accuracy of a measure and the interrelationship and differences among them. The paper ends with a case evaluation of the ethics of a proposal related to an experiment. Primarily, the write-up is considerably general material on the topic but is thoroughly researched about. CORRELATIONAL AND EXPERIMENTAL RESEARCH: CORRELATIONAL RESEARCH: Correlational Research is the type of research which is conducted on the basis of the assumption that “everything affects--and is affected by--everything else”. This assumption further assists that reality can be best described as a network of interacting and mutually causal relationships. The relationships exist in the form of a web, instead of a linear relationship, as in the case of experimental research. These realities create a system and the dynamics of that system (how every single part of the whole system can affect – and does affect – all other parts of the system) is more important than even the causality itself. There are no independent variables; all variables are inter-dependent. The objective is to be able to find or predict the value of one variable, if given the value of another. Here, it is important to mention that the Correlational research is quantitative, and not qualitative. Moreover, the correlational designs do not indicate causality and do not allow one to make causal inferences; since the research cannot determine which variable can cause changes in the other(s). Here arises the “third variable problem”, i.e. both variables mite be being influenced by another unobserved or unnoticed third variable. Furthermore, there might develop the directionality problem, i.e. it might be impossible to understand the direction in which the causality is moving, for example, does aggressive behavior cause people to watch aggressive TV shows, or does watching aggressive TV shows cause aggressive behavior? However, some correlational designs might permit the use of some causal statements; such as path analysis and cross-lagged panel designs. TYPES: Some major types of Correlational Research Designs are: Bivariate Correlation Regression and Prediction Multiple Regression Factor Analysis Correlational Designs Used to Make Causal Conclusions Systems Analysis EXPERIMENTAL RESEARCH: Experimental research is the type of research which is conducted on the basis of the assumption that the causal laws rule the world. These laws are essentially linear, but are comparatively quite complicated and interactive. The main goal behind the conduction of an experimental research is to create cause-and-effect laws, while the main purpose is to eliminate and strike out alternative hypotheses. The more the alternative hypotheses are stricken out, the better the research is. If all the alternative hypotheses are eliminated, it can be said that the independent variable is the cause. TYPES OF VARIABLES: The following variables are taken in consideration while conducting an experimental research: Independent Variables Dependent Variables Control Variables Random Variables Confounding Variables TYPES OF DESIGNS: The main types of designs are: Between-subject designs Within-subject designs Single-factor designs Factorial designs USE OF CORRELATIONAL RESEARCH: Correlational research may be more helpful and efficient if used when: The research is more quantitative than qualitative or completely quantitative. When the laws of causality are not really applicable. When the variables are more interrelated. The research findings are more likely to be web-like instead of linear. VALIDITY, RELIABILITY AND ACCURACY OF A MEASURE: VALIDITY: The term Validity mainly refers to the fact that a property stands true. However, it is important for a statement to be “completely” true in order to be valid, i.e. the probability of it being correct should be 100%, and not less. Such a property, which is 100% true, is called a deductive property, while the property that “might” be true is an inductive property. Thus, in order to be “valid”, a property should be deductive. Validity is also the strength of a measurement; or rather the conclusion or inference. LOGICAL FORM: Logically speaking, validity can be shown as: If all P’s are Q and S is a P, therefore, S is a Q. In order to understand validity better, consider the following statement, which is “invalid”: If all P’s are Q and S is a Q, therefore, S is a P. TYPES: There are four types of validity that are usually taken into consideration for experimentation: Conclusion Validity Internal Validity Construct Validity External Validity There is a large number of threats to each particular type of validity, like testing threats, maturation threats and regression threats etc to the internal validity, while threats like mono-operation bias, confounding constructs and mono method bias etc to the construct validity. RELIABILITY: There are many definitions available and many ways of understanding reliability. Generally speaking, Reliability is the ability of a system or a person to keep on executing it functions and tasks and sustaining the flow no matter how unexpected, ambiguous or difficult the situation and circumstances get. Reliability is the consistency of the measurement, that is, the extent or level to which an instrument or device can measure in the same way when used under the same conditions with/upon the same subjects. Reliability, however, is estimated; never measured. WAYS OF ESTIMATION: There are generally two ways of estimating reliability, they are: Test/Retest Internal Consistency ACCURACY: Accuracy is defined as the degree to which a particular measured value is close to the actual (true) value. Accuracy is often confused with precision, which is the degree to which same results show up with every follow-up measurement. Accuracy is the degree of veracity while precision is the degree of reproducibility. A measurement can be accurate but not precise and vice versa; however, a valid measurement needs to be both; accurate and precise. VALIDITY WITHOUT RELIABILITY: It is possible to have validity without reliability, since it is possible for a measurement to be 100% true, but the same measurement taken again might be different, violating the laws of reliability. RELIABILITY WITHOUT VALIDITY: Similarly, it is possible to have validity without reliability, since it is possible that we get similar results consistently, but the measurement is just not true, or rather valid, in the first place, violating the laws of validity. EXAMPLE(S): We can take the example of EQ (Emotional Intelligence Quotient). When conducting a test for EQ of some people, it is possible that: The measurements are valid but not reliable; like when the people do answer the questions correctly and truthfully, but if the test is conducted again, there is a change in their emotional behavior, which might be totally unconscious, and thus the measurements thus accumulated differ, supporting the laws of validity but violating the laws of reliability. The measurements are reliable but not valid; like when people are not sure and not completely aware of their emotional status and/or try to hide or fake emotions. In this case, the resulting measurements might be consistent (reliable) but not true (valid). Therefore, the conclusions thus accumulated support the laws of reliability but not validity. So basically, reliability and validity do not differ on the mere basis of definitions. Reliability is the degree to which the results are consistent, while validity is the degree to which the results are true. It is ideal to have both validity and reliability, but if only one can be obtained, it is better to have validity than reliability, because if the results and conclusions drawn are not true, it’s no use of them being consistent, at all. EVALUATING THE ETHICS OF A PROPOSAL: Before conducting any such experiment, the ethical implications and guidelines should be kept in mind, like: The participants should be fully aware of their participation. The participants should be agreeable and willing towards the participation. Their personal spaces should not be violated. There should be no harm to the participants in terms of the therapy. Incase if either one of the group is noticed to be harmed by the respective type of therapy (face-to-face or videoconferencing) the research should immediately be stopped. Since videoconferencing involves the use of the internet, the ethical implications of using such means are applicable here as well like the privacy of the participants, the protection of their rights and security of the personal information of each participant, etc Works Cited 1. Anthony E. Kelly, Richard A. Lesh. Handbook of Research Design in Mathematics and Science Education. Lawrence Erlbaum Associates. 2. Joseph Alex Maxwell.(2005). Qualitative Research Design: An Interactive Approach. Sage Publications. 3. Marty Sapp. (2006). Basic Psychological Measurement, Research Designs, And Statistics Without Math. Charles C. Thomas Publishers. 4. Paul E. Spector. (1981). Research Designs. Sage Publications. Read More