Validity: External, Internal, Construct, Essay Example

In research paradigm, validity and reliability are the most basic characteristic issues used in qualitative and quantitative analysis. Validity as a psychometric standard is embedded in a positivist approach, which is relevant in reflecting on the qualitative point of view ascribed to the establishment of the truth. In view of this, definition of positivism ascribe to a theory systematic to validity. Additionally, other empirical conceptions culminating from resided validity include truth, deduction, universal laws, evidence, reason, and actuality among others.

For any qualitative research conducted, validity as a concept is a determinant that measures the truth concerning the investigation. Validity is measured by researchers by asking questions and answers are retrieved from other people’s research work (Joppe, 2000, p. 1). Other researchers define validity in quantitative research as construct validity. Construct implies initial concept, hypothesis, notion or question determining data gathered and method of data collection. However, quantitative practitioners affirm the cause or effect interplay between data and construct for validation of investigation by applying test procedures or processes (Golafshani, 2003, p. 599). As a result, with regard to validity, researchers conclude that, it is whether measurements of the mean are accurate or they are measuring the intended features. Accuracy of the mean helps in relating the cause-and-effect relationship present in internal validity.

The above definition is associated with quantitative research methodology. It summarizes that validity to be the extent in which instruments measure the exact thing it purports to measure. An example of validity in research is when screening of applicants for job positions, tests are applied, and scores relate directly future performance of the job by employees. This can be in relation to the construction industry and project managers use this approach to test their performance (Golafshani, 2003, p. 610). This helps managers and leaders improve on the weak areas and gain competitive advantage in the industry. High performance of projects in management leads to execution of many research proposals that will positively affect the industry and increase productivity. This makes it easier for investigations to be performed since, accurate information is reached and obtained in a short period while utilizing theories relating to the field of study. Validity is classified into different categories like content, face, construct, concurrent, factorial and criterion-related validity.

Construct validity is the degree to which measurement of instruments to aspects like trait and theoretical framework construct are under measurement. In theory, construct validity is an ongoing process that enables individuals predict test scores in diversified situations and settings. This as an assessment of well translated ideas and theories into measures and actual programs (Trochim & Donnelly, 2008, p. 57). This validity is measured by using two distinct perspectives, which are the views of definitionalist and relationalist. Reliability and validity are two concepts that work together in ensuring that measures intended are reached. Problem solving techniques and question answering motivates employees and trains employees; therefore, project managers need to test the validity of their research by responding to questions stipulated during the research work.

The threats related to construct validity is inadequate preoperational explication (meaning) of constructs. Preoperational means those translations of constructs measurements or treatments made before research projects (Trochim & Donnelly, 2008, p. 8). Mono-operation bias is also a problem and relates to independent variables, treatment (program), cause in the course of investigation and, therefore, does not pertain to outcomes of the research. Different treatment interactions resulting from a combination of programs separate from participation of individual research project management conducted by others. Mono-method bias also contributes to the threat of measures and observations, not of programs or causes. Social threats of construct validity include hypothesis guessing, expectancies of experiments and apprehension evaluation. Guessing of the hypothesis is when participants base their behavior on their guess of the study and not on treatment or observation accorded to the study. In a population, many people become anxious on evaluation and thus, the results performed are either good or poor. This hinders the reliability and validity of the project since results achieved do not replicate the ones achieved by previous researchers (Trochim & Donnelly, 2008, p. 10). Participants give answers desired by researchers and thus, the response is wrong to be labeled as treatment effect since the end does not meet the means of the study. In this scenario, participants’ reaction aims at looking good for their researchers and then, it gives wrong conclusions and recommendations of the study.

For the paradigm of qualitative research, validity does not apply to this methodology, but their concerns include aspects of precision, transferability and credibility. Content analysis validity in research is achieved through the use of multiple classifiers that enable individuals arrive at an agreed category definition. However, this analysis faces challenging situations linked to reaching a conclusion with inferential procedures. In effect, conclusions reached needs to have supporting evidence from collected data or have explanations supported by other phenomena. Generalizability (external validity) of research information is reached at when validity (trustworthiness) is maximized or testable to achieve results that are credible and defensible. Although generalizability gives clear distinctions on quantitative and qualitative research methods, it is a criterion that looks at the quality of case studies depending on selected and studied cases (Trochim & Donnelly, 2008, p. 58).

Validity and reliability are dependent concepts that determine the impact of research project or findings interpreted from previous research work. Validity is viewed to derive questions that guide the researcher on how to determine the process of measurement, assessment or project management measured. That is, the actual thing the research intends to measure. For construction project management, it is vital for the researcher to determine the trustworthiness of the project to be conducted and assesses the measures used in order to succeed (Trochim & Donnelly, 2008, p. 59). When repeating variables and measurements in research projects, consistent results need to be obtained and thus, supports reliability concept in research.

Validity has two main parts in research projects; internal and external. Internal validity, for instance mean difference between control groups and treatment, covers results of research study conducted by looking at whether there is legitimacy. This is important because of the way the selection of groups was conducted, recording of data and performed analysis of the case studied. For construction project management, poor internal validity is achieved when performed tests use different treatment and control groups in displaying findings. Also, if perplexing variables are not accounted for during the design construction stage of project management or analysis process.

A discussed earlier on, generalizability is also known as external validity. It is the degree to which study conclusions hold for other people in other places and times. In essence, this involves whether studied results of a given project are transferable to others interest groups (populations) like racial, gender and other geographic subgroups. Internal validity goes together with external validity and this contributes to the success of the projects since poor results obtained are not applicable in population designs. Internal validity is threatened by reliability and researchers should always be vigilant on how to go about this issue. This is in order to reduce risks involved in assessing data presented during the study. For instance, dictations of research protocols give subjects weight measurements, and scales provided need to match with replicated studies taken for same periods. External validity is threatened by wrong explanations when making a generalization, the time, people and place setting.

External validity can be improved by using random sampling models to draw conclusions from a given population. Proximal similarity theory should be used effectively in reducing external validity in that, study conducted should be done in a variety of places, with diverse people and at different times (Scandura & Williams, 2000, p. 1248). Threats to validity can be conceptualized and controlled through the use of selection bias, which expands or purifies on the group of experiment. Generalizability of findings can be viewed using simple designs as portions of potentially large designs. Also, external and construct validity replicates hetero-method. The improvement of external validity contributes greatly to the reliability of research and data collected with a view of conducting further research in laboratories to test their accuracy.

The impact of validity on research methodologies applied will be determined by topic examined and method employed in examining generalizability and robust information collected. Research can be impacted both negatively and positively depending on the degree of their consideration the internal and external validity threats in order to incorporate them into the conclusions and recommendations to be effected for future methodologies contributing to future research methods employed in the study (Creswell, 2000, p. 125). Researchers need to put into measurements that need to be validated in the research study. Thus, making conclusions that will be given for future research to be conducted. For construct project management, the validity of the research will be dependent on the variables used during the research and variables employed in giving qualitative and quantitative findings. Qualitative and quantitative research methodologies give different meanings to validity concepts, but researchers need to ensure that method used during data collected is trustworthy for findings reached to be appropriate and viable for use during future research (Creswell, 2000, p. 128). Research studies impact validity by answering questions during the project work and giving substantial issues that emphasize on the validity and reliability of the study.

Validity in research projects needs to be given emphasis in order to increase future research in management disciplines (Scandura & Williams, 2000, p. 1250). Internal, external and construct validity are similar because individuals are able to draw conclusions from well-informed research central to data collected. Trochim gives a detailed explanation of construct validity in chapter three. He refers to it to mean operationalizations, where ideas or construct translation into something real occurs. Consequently, when individuals present a program to create, the concept of construct validity will examine the degree to which, reflection of actual program to the ideal program has been envisioned in the research (Trochim & Donnelly, 2008, p. 56). The existing characteristic ascribed to construct validity is used when programs characteristic are measured. The similarity of internal, external and construct validity is the fact that they are measurements used in research methodologies in presenting information of studies conducted. These validities are important in providing robust information to the population on research and how they impact the future of research. It is imperative for all researchers to be able to identify the variables affecting all validity aspects. This is because it gives them a direction on how measurement of the variable will be attained. Instrumentation, testing, maturation and history are the most common variables related to external validity (Creswell, 2000, p. 126).

In conclusion, external validity in research paradigms accepts universal theoretical measures and rejects conceptions of common sense like random sampling automatic superiority. It is evident that researchers are not able to emphasize inferences of external (generalizability) validity at a point in time hence; interpretation should not be made to assume that researchers have ignored how theoretical progress is reached. Also, it should not be concluded that possible improvements can be reached. Internal validity emphasizes on the cause-and-effect relationship (causality). External validity concerns generalization of research study across settings, individuals and time. Construct validity looks into how well employed measures fit into theoretical frameworks presented in testing designed project management issues. The impact of validity goes a long way in determining the measures, data collected assessment and information provided during the course of study. In the case of project construction management, leaders need to be aware of the variables affecting internal and external validity to be able to convince the readers on aspects introduced and needing further research.

References

Creswell, J. (2000). Determining validity in qualitative inquiry. Theory into Practice, vol. 39, no. 3, pp. 124-131.

Cozby, P & Bates, C. (2012). Methods in behavioral research. New York, NY: McGraw-Hill.

Golafshani, N. (2003). Understanding Reliability and Validity in Qualitative Research. The Qualitative Report 8 (4), 597-607.

Scandura, T & Williams, E. (2000). Research Methodology in Management: Current Practices, Trends and Implications for Future Research. Academy of Management Journal 43 (6), 1248-1264.

Trochim, W., & Donnelly, J. (2008). Chapters 1, 3, 5, 6, 7