ANDA'S IT LIBRARY
DATA PROJECTS

OVERVIEW
Anda Vitols specializes in
  1. Psychomentrics
  2. Judgement Error Analysis
  3. Modeling Error Analysis
  4. Foundational Psych-Social Demographic Analysis
Specifically, Psychometrics, are personality inventories that can be used in targeted marketing to change client behaviour. Data Error Analysis uses statistical and research methods to determine the reliability and validity of results. In real life, Judgement Error Analysis use simulations to measure variabilty in an organization's decisions. Modeling Error Analysis looks at data point validity. It measures the truthfullness of model representation.
  • Data Exploration, Modeling and Mining
  • Behavioural Statistics
  • R's Machine Learning Stack
  • Explainable Artificial Intelligence (XAI)
WHY IMPORTANT
HOW IT’S DONE
Psychometrics - using social media platforms
  • To create personality inventories (that mirror academic inventories in structure)
  • Then create targeted marketing to change behaviour.
Judgement Error Analysis - using simulations
  • To measure human decision-makers' variability in results
  • Then create better procedures to reduce variability
  • This helps corporations find the sweet spot in client offerings, saving them significant money; or helps institutions increase fairness on applicant submissions.
Modeling Error Analysis - using behavioural research standards
  • To measure truthfulness of machine learning model representation
  • Then create or find data points to improve model validity - that the model is indeed describing what we want it to describe.
DETAILS
PSYCHOMETRICS: for targeted marketing to change behaviours.
Step 1) Use social media platforms to create personality inventories.
Step 2) Cross-reference with demographic and attitudinal data points.
Step 3) Create content using these data points - that actions change in behaviour.
Step 4) Use social media platforms to micro-target this content.
Performance Measurement and Transparency:
  • Because on social media, the effect is immediately known. We can change content on the fly to improve results.
Personal note - caution:
  • Please note - there are actors weaponizing this powerful technology, by pushing false narratives, trick narratives, emotional and physical threats, etc. This is not necessary! It speaks more to these actors' state of mind, where it appears they get a "dopamine rush" by causing destruction. However, it is necessary to use this technology to counter these actors, by creating an even playing field.
  • Normal healthy people get a "dopamine rush" when creating projects of growth and abundance, for the highest good of all. When kind, caring people come across judgemental conversations, or people behaving erratically, they create a safe space for these people, to find out what is really going on... or for themselves when mistakes are made or weirdness is happening.
    Please note - this technology was initially designed to be used by normal people - people who are kind, who take responsibility, and take action from that space.

DATA ANALYSIS in R
Analyses by Anda Vitols
ABOUT PSYCHOLOGICAL TRAIT INVENTORY SCALES
Trait Analysis: The Big 5 Personality Inventory
The science behind creation of the 'Big 5 Personality Inventory'.
Exploring the personality trait, "Openness to Experience".
  • Modeling the 'Openness to Experience' Trait (2023-2024)
  • Let's apply the personality inventory, to further test the delineation of the Big 5 categories. We look at one of the personality traits in the dataset (openness to experience), and see how the other traits compare to it.
  • Procedures: logical regression / decision tree / knn / random forest
Compare Traits to Demographics
Comparing personality traits to socio-economic information from geographic regions.
  • Analysis of 4 Psychological Regions of the United States (2023-2024)
  • Now let's get real. Let's compare some regional demographic information to psychological profiles of different regions that have been collected from previous studies.
  • Procedures: contingencies / correlations / descriptives / frequencies / one factor ANOVA / linear & quantile regressions / hierarchical clustering
Blogs about Trait Analyses
HEXACO (big 6)
  • "Big Five" is a trait analysis tool used with big data for marketing and influence platforms. HEXACO adds another dimension to describe our traits. This dimension checks a person's motive to lie. This successfully addresses the weakest part of the original Big Five Scale.
  • How to Spot a Dishonest Person: The role of honesty-humility in personality (2023)
  • How Honest Are You About Your Religion?: A summary of chapter 8 of 'The H-Factor of Personality' (2023)

ABOUT SOCIO-ECONOMIC DEMOGRAPHIC PATTERNS
How Happy are People?
Using surveys to collect psychological data.

Do Elite Athletes Cluster Together?
Collecting demographic data from sanctioned data posted on the web.

How are People Using the Internet?
Using Google's search engine and 'Google Correlate' to collect trend and demographic data for analysis.

Data for Good
About 'Data for Good'
House of Friendship Datathon
  • To help House of Friendship with their community mission, we looked for patterns in demographics between their volunteers and donors.
  • House of Friendship Datathon (datathon 2019 - my report 2022)

DATA EXAMPLES in R
Reference Documents by Anda Vitols
Examples 2023
examples - sections 1, 2, 3 (2023) prepare data | describe data set | data modeling
examples - section 4 (2023-2024) data mining
  • Exploring tidyverse, piping and other tools that have become more intuitive and efficient for wrangling tasks. Delving deeper into modeling and mining. R is now fully integrated with non-proprietory apps, as well as having a whole set of publishing apps for document and application creation.
Examples 2019
examples 2019
  • One, two, and multivariate examples of continuous and discrete dimensions. Data wrangling. (2019)
  • Quick-look inventory of datasets, packages and special functions used for each example. (2019)
THIS & THAT
Some Outdated Blogs
COVID-19 Resources for Data Scientists (report: 2020)
How R fits into Corportate Data Projects (2019)

back home