Estatística Psicobio I 2026 #01 - Introdução à Teoria da Medida, Variáveis, Fatores

Main ideas / concepts taught

1. Purpose of the course

   • Introduces Statistics Applied to Psychobiology (postgraduate level).
   • Emphasizes a cumulative “staircase” structure: each class builds on the previous ones with small steps.
   • Uses an approach combining quantitative methods first, then moving toward qualitative later (Courses 1–3 include methods broadly).

2. Modern definition of science (foundation of the course)

   • Science is defined as:
     • the use of the scientific method by an agent that performs self-observation.
   • Contrasts with an older, narrower definition (“science = scientific method”), which ignores the agent and self-observation.

3. Education vs. science vs. technology

   • Education (as commonly practiced):
     • training skills whose “ultimate goal” is to produce a social actor—typically shaped to fit the labor/work structure.
   • The course argues that scientific education (in the true sense) is largely missing in real life/institutions.
   • Technology:
     • the mere application of the scientific method/tooling (e.g., running statistical procedures to produce results).
   • Scientific education (the course’s aim):
     • learning tools so you can use them for self-observation—to change how you think/act, not just compute outputs.

4. Goal for students’ internal movement

   • The course is meant to produce productive anguish (not “happiness”).
   • Distinguishes:
     • Anxiety: a negative state facing an inescapable situation; tends to cause avoidance.
     • Anguish: a state arising from uncertainty with a course of action (a path), increasing probability through effort.
   • The course aims to reduce unproductive anxiety by giving tools and structured questioning.

5. A method-level view of statistics (how to “do science”)

   • Core cycle taught early:
     1. Start from a phenomenon.
     2. Decompose it into attributes/variables.
     3. Define how these attributes form an abstract construct (factor).
     4. Ensure measurement “rulers” are valid and precise.
     5. Use statistics to build a statistic of interest, then model uncertainty via distributions.
     6. Perform hypothesis testing to answer research questions.
   • Emphasis: if statistics are used instrumentally only (button-pushing), it becomes technology, not science.

---

Detailed methodology / instruction-like content (step-by-step)

A) “From phenomenon to hypothesis testing” pipeline (conceptual workflow)

• Step 1: Define the phenomenon

  • Examples used: measuring ages, and later examples like flower or depression (mental health context).

• Step 2: Decompose the phenomenon into attributes

  • Identify measurable attributes (e.g., for a flower: size, color, shape).

• Step 3: Build descriptive measurements

  • Compute descriptive statistics to summarize data (examples mentioned):
    • mean
    • variance
    • standard deviation
    • standard error
    • confidence interval
    • coefficient of variation

• Step 4: Choose a “statistic of interest”

  • Based on the research question, define the statistic that represents what you want to learn.

• Step 5: Understand the distribution of that statistic

  • Use the distribution of the statistic of interest to quantify uncertainty.

• Step 6: Perform hypothesis testing

  • Use the hypothesis test as the tool to answer the research question.
  • Example scenario:
    • compare groups (e.g., medicine vs. no medicine) and test whether their distributions differ statistically.

• Step 7: Use results for self-observation

  • The “science” part is using what you learn to justify/adjust your own behavior as an agent, not merely to publish.

---

B) Measurement theory “ruler” requirements

• Before collecting data, evaluate the measurement instruments (“rulers”).

  • A ruler can be: a test, questionnaire, blood exam, interview, etc.

• Each ruler must have two key properties:

  • Validity
    • It measures what you intend to measure (matches the concept/attribute).
  • Precision
    • It has sufficient granularity to distinguish relevant differences.

• If either fails:

  • the measurements can be systematically wrong (e.g., underestimating/overestimating prevalence).

---

C) Variable and factor handling rules

• Variable (definition)

  • “Any data directly observable and quantifiable.”
  • Variables are tied to attributes, not the full phenomenon.

• Factor (definition)

  • A latent/grouped construct not directly observable on its own.
  • Built as a combination of multiple variables (often via factor analysis).

• Important caution:

  • Don’t treat a factor as if it were directly measurable with a single question.
  • Instead:
    • identify directly observable variables tied to the phenomenon,
    • then combine them using a function/model to estimate the factor.
  • The course hints at using factor analysis to estimate weights connecting variables to factors.

---

D) Example-driven measurement logic (how invalid/low-precision instruments distort conclusions)

• Depression prevalence in dialysis patients (questionnaire example)
  • If a questionnaire is valid for the general population but lacks items relevant to dialysis-specific experiences (missing relevant attributes/questions), then:
    • it will score dialysis patients incorrectly (often lower),
    • leading to underestimated prevalence.
  • Framed as failure of validity and/or precision of the ruler for that phenomenon/population.

---

“Core lessons” emphasized repeatedly

• Statistics isn’t just computation

  • The crucial question is what you measured and whether the measurement is valid/precise.

• Psychology risk

  • Students may learn attributes/labels without understanding the phenomenon-measurement relationship.
  • The critique includes simplistic “trait lists” presented online as if they directly measure internal states.

• Research ethics

  • Ethics in research is framed as training + scientific rigor, not primarily “kindness/affection.”

• Academic responsibility

  • Advisors/supervisors may not fully care about every research question; the course stresses student responsibility and commitment to proper research conduct.

---

Speakers / sources featured (identified in subtitles)

• Altaí de Souza (speaker; researcher at UNIFESP Department of Psychobiology)
• Professor Maria Lúcia Formigone (Professor Malu; co-teaches)
• Zé (course monitor)
• Pedro (Pedro Zangrano and Pedro Alvim mentioned)
• Gilmara (student/questioner)
• Marcos / Marco (student/commenter)
• Letícia (student/commenter)
• Rafael (student/commenter)
• Amanda / Grasiela (student/commenters)
• Marcelo (student/commenter)
• Márquez / Gustavo (name mentioned as participant/commenter)
• Rilo (mentioned in an illustrative comparison)
• Jung (Carl Jung; referenced as an idea about becoming “not what you were”)
• Lacan, Freud, Skinner (referenced as authors; said not strictly required here)
• Lakatos (I. Lakatos referenced for “research project” definition; also “Inri Lacatos” appears as a subtitle error)
• Malinowski (referenced in the context of a movement described in Course 3)
• Spirmen (subtitle appears as “Spirma 1916”; likely Spearman (1916) referenced for “general intelligence”)