3.11 Validity and Reliability Of Research

when you are writing a research you need to be aware of two terms your research needs to be reliable and valid and in this video I would like to discuss these two terms to you because they are very important it is I have seen it many times that a student who is writing the degree thesis comes to a seminar where he is supposed to defend his thesis and now the examiners ask to all questions why do you think that your research is reliable and why do you think your research is valid so let's examine what is hidden behind the reliability and validity of a research and we have a definition of a reliability here reliability refers to whether your data collection techniques and analytic procedures would reproduce consistent findings if they were repeated on another occasion or if they were repeated by another researcher this means to have a reliable research meaning that if someone else have decided to select similar sample as you and let's say similar data collection techniques similar analytical techniques to evaluate the data if this other person on a different time would produce a similar findings to yours so the idea is simple and now there are certain biases and errors that can influence the reliability of our research so let's look at them first of all there is the participant error this is any factor which adversely outers the way in which a participant performs let's say that we are doing and interviews with our participants now we are deciding about the time of the day when we are interviewing this our respondents well for instance if we would select to interview them at 11:00 in the morning they can produce different answers than if we would interview them at 1:00 in the afternoon because they are going to be after lunch they are going to be more calm maybe or happy so be careful so that no participant error occurs in your research secondly we have the participant bias this is any factory which produces a false response let's stick to our example that we are interviewing employees with in some organization let's say we would decide to interview them in these open offices they are very common nowadays where a lot of people are sharing one office and we would be interviewing your participant in the middle of such open office this participant can be afraid that his answers will be overheard by someone else so the participant may be sort of falsely to positive with his answers because of this being afraid of being overheard by his manager or someone else so again be aware of this then we have a researcher error this is any factor which alters the researchers interpretation for instance if you are doing again interviews and you plan all of your interviews but say you are doing aid of them you plan all of them within one day then such researcher error can occur because after conducting six interviews you are going to be very tired and in the last two interviews you may not be asking the questions in a proper way so maybe a better choice would be to help two interviews every day throughout four days finally we have the researcher bias and this is the most common one this is any factor which induces bias in the researchers recording of responses I mean we as the researchers are simply a human beings and as a human beings we are being subjective so we may interpret the results that we got from our respondent in a subjective way so be aware of it and try to stay objective when you are doing this for instance interviews now when you are on a seminar while defending your thesis asked about why do you think your research is reliable you should address these four points these two errors and two biases and justify why they did not occur in your research so that your research is reliable now we come to the invalidity of a research and we have three different kinds of validity we have construct validity internal validity and external validity so let's start with the construct one this is it may be the simplest one the construct validity is concerned with the extent to which your research measures what it claims to measure behind this fancy sentence are all of these methodological choices that we have made while designing our research so the justification of why we selected this research nature this particular research strategy and so on and so on now let's be more practical if you are asked on a seminar while defending your research why do you believe that your research is achieving the construct value B your answer should have sort of a top-down approach so you would be like so okay I wanted to research the IT skills of employees this was my research idea to do it no one before has done it so I had to select exploratory research nature that makes sense right I had to be the first one to explore this phenomena now exploratory research nature implied then inductive research approach so you see these two work together then inductive research approach sort of again implied quantitative methods and you continue this way so from the top or the most general things you justified to the most concrete ones and that way you will really prove that your research is achieving the construct validity second of all we have the internal validity internal validity is established when your research demonstrates a causal relationship between two very and this is especially true if our research strategy is experiment because in experiment we clearly have to define the independent variable and dependent variable and then we are examining the relationship between the store so what internal validity is asking about is why are you making this assumption from the early beginning that this relationship is existent so let's say that we choose as an independent variable the educational level of our customer support it these guys who are answering the phone when the customer has a problem and as a dependent variable we select customer satisfaction level so our assumption is that the more educated the support employees will be then the happier the customers will be and the more satisfied they will be what is internal validity asking about is why are you making this assumption why do you believe that this relationship between these two variables exist so if you are asked about the internal validity this is the point that you should justify finally we have the external validity external validity is concerned with whether a studies research findings can be generalized to other relevant settings or groups we are going to discuss this more in the later modules when we come to the data collection techniques and especially sampling this is about sampling when you are making a research you are choosing a sample from a population because you simply cannot research the wall' population you have to select most likely some sample what external validity is asking about is whether when you research this sample and you make some findings from this sample whether the these findings will be generalizable to the wall population and the it is essentially asking about the represent representativeness it's hard word to pronounce whether your sample is really representative for instance you are having a mall poor case study and now you're selecting ten companies from one particular field now the question is have you selected a successful companies or unsuccessful companies have you selected a mixture of successful and unsuccessful companies to have a more representative sample so you see when you are asked about the external validity you should justify the choice of your sample and that's all and yet now you will have all these points justified you will say why your research is reliable why this valid from the construct perspective why is it valid from the internal perspective and finally why is it valid from the external perspective

Channel: MeanThat & Authentic Data Science

Views: 260,592

Rating: 4.8540449 out of 5

Keywords: research, research thesis, writing a thesis, writing academic paper, writing academic research, research design, research approach, literature review, research strategy, primary and secondary data, sampling of research, validity of research, reliability of research, validity and reliability of research

Id: 2fK1ClycBTM

Length: 9min 20sec (560 seconds)

Published: Thu Mar 17 2016