Internal Validity looks at the dependent variable and whether or not changes to the dependent variable are caused by the independent variable. In field survey research, the internal validity is low because of the inability to manipulate the causal variable, making it harder to determine the cause of the changes to the dependent variable.
Issues that impact validity:
Extraneous variables can negatively impact internal validity, unless they can be controlled
Respondent bias can occur when the participants do not answer the questions truthfully, but rather how they think the questions should be answered, which can impact the results of a survey and decrease the internal validity
External Validity refers to the generalizability of the results, and how the results can apply to the broader population and in different contexts. Survey research can have greater external validity and results are more generalizable than stricter experimental type research. Often research that has greater external validity will have lower internal validity, however that is not always true as longitudinal surveys can be strong in both external and internal validity. For surveys, external validity can be strengthened by randomization of the sample.
Case Study
In the case study, the external validity would be high, making the results more generalizable. This generalizable source of information could be shared with other teams, tested with other populations and implemented into practice if the model turned out to make a positive change.
Construct Validity has to do with how well the tool is measuring the construct that is being researched. If the survey is measuring satisfaction with a new team oriented model of care, the construct of satisfaction must be clearly defined and measured with the tool chosen. Satisfaction could include ideas such as do the staff have an increase in morale with the new model? Is the staff generally happier at work, are there less incidents occurring that are being escalated?
Statistical Conclusion Validity implies that the correct method for processing the data was used to test the hypothesis. This method would also need to meet the requirements of the sample size, and the concepts researched in order to be congruent with the research.
In terms of the case study, ensuring the sample size was sufficient and the correct methods were used to process the data would strengthen the statistical conclusion validity.