SES - Create standardized Higher Consciousness composite variable

[lawof1]

Analysis Techniques

An initial model needs to be built to begin quantifying the degree to which someone has experienced Higher Consciousness. Analysis techniques may include

• Pairwise correlations analysis of variables.
• Factor analysis

To begin, we first need to understand the structure of the data identify unique types of experiences based on correlated variables. In the simplest form, this can be done by pairwise correlations analysis of variables. However, similar questions can also be related to each other and define unique experiences. To identify the underlying structure of the data, Principal Axis Factoring with Promax rotation is used since its results are easy to interpret and is often used in psychometric research.

Variable Selection

A number of post-analysis concerns arise around variable selection for a meaningful composite variable:

• Predictive power of the variables to identify the condition of Higher Consciousness.
• Eliminating redundancy
• Properly weighting variables

For example, using Item Response Theory and Factor Analysis to identify key variables and then combining them using an averaging technique.

The process of selecting variables needs to be done with these criteria in mind. To validate the composite variable:

1. Create an initial Higher Consciousness (HC) composite variable (model).
2. Conduct interviews and case studies with individuals and add a subjective rating to the data.
3. Validate and improve the HC model.

Ultimately, the HC model should be designed to correctly identify and predict genuine experiences of Higher Consciousness.

Composite Variable Creation

As an example, a composite variable can be created through one of the following techniques

• Averaging / summation - Find the average of the Likert questions for Higher Consciousness.
• Factor analysis scoring - Identify a factor which represents Higher Consciousness and use the per-person factor scores.

Next Steps

Next steps include finding resources such as the one below to validate our approach in creating an initial HC model. Analysis should be run and variables identified. Separate analysis should be conducted, and then results interpreted and then discussed. Finally, an approach should be agreed on and the HC variable built and operationalized in R for the SES and PBS surveys.

Which method should I use to present the Mean of a 5-point Likert scale?

[lawof1]

Following up from the above, here's what needs to be programmed in R:

• [ ] Generate factor scores for all factors (mystical to energy)
• [ ] Generate correlations matrix for the above factor scores
• [ ] Generate correlations matrix on "consciousness" factor against all individual questions (mystical to energy)
• [ ] Identify groups of questions and create an initial model, e.g., 3 consciousness questions, 2 energy questions, 2 bliss questions. Create by summing and averaging.

This is designed to be an initial model that will later need validation and improvement. It is meant to serve as a starting point for future analysis only.

@DivineKali @LazC2G

[lawof1]

Some IRT information that might help in variable selection. The psych packages fa function may provide some additional IRT output, specifically item difficulty, which might be a data point in helping in variable selection:

https://www.metheval.uni-jena.de/irt/VisualIRT.pdf

[DivineKali]

Do the R programming work in ses.R