How would you address discrepancies in data collected from different sources in your research study?

Discrepancies in data collected using the same tools and techniques indicate that either demographics or any other aspect of the sample may be responsible for difference in the data, and hence this aspect needs to be either controlled or considered while making generalisations. This is an important aspect of research and shall not be ignored.

My first approach of addressing data discrepancies from different sources would be to see it in a positive light: to be certain your overall research is valid you need to conduct it in a variety of representative social or natural environments. As example, if your data should show how the health and happiness of the residents in a country, the results will be different depending on whether wealthier or poorer people get asked. It can advance your research if you put different data results into such subgroups. Even in strict clinical experiments data results can vary, some relate to language use and human factors, some point out that the research needs to get better structured and delivered next time.

Two seemingly identical studies can both be correct and yield very different answers because the real world has a lot of variation. Of course, its important to check for outright errors. However, too often anomalies are dismissed as an "error" or "superfluous", when they really highlight a breakthrough or significant outcome.

Address discrepancies in data collected from different sources by: Verify Sources: Ensure data sources are reliable and consistent. Normalize Data: Use standardized methods to align different datasets. Cross-Check: Compare data points across sources to identify discrepancies. Resolve Differences: Investigate and reconcile discrepancies through additional research or adjustments. Document Process: Maintain a transparent record of data handling and resolution methods.

Looking at the data in respect of the context is crucial. I have collected data using interviews, and the themes have been entirely different from the ones done with the same population (target) but a different sample (by other researchers previously), and this helped me realise that the context matters, the situations and challenges in the life of the participants matter, specially in my area of study- psychology. Hence, variances in finings allow us to have a broader idea about the participants, which is helpful.

Understanding data context: - Consider historical, cultural, and demographic factors influencing data. - Recognize discrepancies: Economic shifts, inflation, or market changes affect data alignment. - Adjust interpretation: Analyze context to accurately interpret data. - Detective work: Piece clues for a coherent understanding and informed decision-making.

Documenting the similarities and differences both, are what makes a research a good and an ethical research. One may choose to also explore the causes and factors which may be responsible for differences in the findings, and this becomes yet another finding of the research.