Introduction
Research strategies determine how mixed methods research will be designed in studies involving both quantitative and qualitative data, establishing the order and emphasis of data collection, study design, and analysis sequence. Mixed-method strategies can be broadly classified into sequential and concurrent approaches (Creswell & Creswell, 2018).
Sequential Strategies
Sequential strategies involve collecting and analyzing one type of data first (either qualitative or quantitative), followed by the other, with the second phase building upon the first.
Sequential Explanatory
Sequential explanatory follows a quantitative to qualitative sequence design. The study starts with numerical data collection and analysis, followed by in-depth qualitative examination to explain or expand on the numerical results.
Sequential Exploratory
By contrast, sequential exploratory follows a qualitative to quantitative design. The study begins with qualitative analysis, followed by numerical study to provide further strength or validity to the qualitative insights.
Sequential Transformative
Sequential Transformative can use a qualitative-first or quantitative-first approach. However, the design is guided by a specific theoretical framework (e.g., disability studies, feminist movement, or critical race theory). The design is often used in studies designed to influence social policy or elicit transformative change.
Concurrent Strategies
Concurrent strategies involve collecting and analysing data simultaneously, rather than in sequential phases.
Concurrent Triangulation
Both qualitative and quantitative data are collected simultaneously and analysed separately. The results are compared to cross-validate the findings, providing stronger conclusions due to the enrichment of perspective provided by different research methods.
Concurrent Nested
Data collection and analysis is prioritised with qualitative or quantitative methods, whilst the other plays a supporting role with secondary contributions.
Concurrent Transformative
Both types of data and analysis are conducted simultaneously, within the framework of a theoretical lens or perspective. The study is particularly useful to advocate for cultural change, challenge social injustices, and influence policy improvements.
Application in Value Investing AI Models
In designing a machine learning model replicating value investing principles, a concurrent triangulation design is best suited, following the multifaceted nature of value investing analysis. Both qualitative and quantitative factors play a decisive role in stock selection, with potential asset acquisitions declined if one facet of the value investing analysis does not meet expectations, regardless of how well the other aspects may fare.
For example, a business may produce outstanding performance metrics, but purchases may not be recommended if they occur during a period of stock market exuberance conducive towards overinflated valuations. Additionally, quantitative and qualitative findings may support each other. For example, lacklustre business performance may be correlated with the lack of a competitive advantage, with both distinct findings supporting the same conclusion of not purchasing the stock.
Due to the complexities of value investing in unpredictable stock market landscapes, gaining the richness and multi-faceted insights of concurrent triangulation will offer the higher chances of success in designing a value investing model suitable for real-life deployment.
References
Creswell, J. W., & Creswell, J. D. (2018). Research design: Qualitative, quantitative, and mixed methods approaches (5th ed.). Sage publications.
← Back to Blog