Video summary

🔍👉 TAMAÑO DE MUESTRA PASO a PASO 👍

Main summary

Key takeaways

Educational

Main Ideas and Concepts

  • Sample size in research is crucial: the video explains, step by step, how to determine the number of subjects/elements needed for a study.
  • What a sample is: a subset of a population intended to be representative, usually because it incorporates:
    • a confidence level
    • a margin of error

Why Sample Size Matters (Benefits)

  • reduces costs (money, labor, materials)
  • allows data collection in less time
  • makes it possible to study very large populations

Types of Populations

  • Finite population (limited number of elements), e.g.:
    • all inhabitants of a community
    • students in a specific university
    • workers in a company
    • fleet of vehicles in a transport company
  • Infinite population (extremely large; cannot practically count all elements), e.g.:
    • insects in the world
    • stars in the universe
    • grains of sand
    • water drops in a lake

Key Parameters Used in Sample Size Determination

  • Sample size (what you’re solving for): the number of elements extracted from the population.
  • Population/universe size (N): the total number of elements in which the studied characteristics may exist.
  • Confidence level
    • represents the degree of certainty the researcher wants (as a percentage)
    • linked to a statistical value (z-value / “zeta value” in the subtitles), using tables
    • examples given:
      • 95% confidence → z ≈ 1.96 (subtitles show “196”)
      • 90% confidence → z ≈ 1.645 (subtitles show “1645”)
      • 99% confidence → z ≈ 2.58 (subtitles show “258”)
  • Margin of error (E)
    • the maximum accepted estimation error due to random sampling
    • example: if a survey reports 60% with margin of error 5%, the true value is expected to fall between 55% and 65%
    • chosen by the researcher based on desired certainty/criteria
  • p and q (probabilities for the event)
    • p = probability the event of interest occurs (probability of “success” / expected proportion)
    • q = 1 − p = probability the event does not occur (probability of “failure”)
    • examples:
      • if p = 30%, then q = 70%
      • if p is unknown, they recommend setting p = 50% and q = 50% (equal likelihood of success and failure)

Methodology / Step-by-Step Instructions (As Presented)

  1. Step 1: Identify the population type

    • Determine whether the population is finite or infinite and use the appropriate population size context (N for finite; conceptually unbounded for infinite).

  2. Step 2: Define the variables needed for the sample size formula

    • Determine:
      • N (population/universe size, if finite)
      • confidence level (chosen by the researcher)
      • margin of error (E) (chosen by the researcher)
      • p (probability of the event occurring)
        • usually taken from past research
        • if unknown → p = 0.5
      • q = 1 − p (probability of event not occurring)

  3. Step 3: Use confidence level tables to obtain the z-value

    • Look up the zeta/z (z-score) corresponding to the chosen confidence level.

  4. Step 4: Substitute parameters into the sample size formula

    • The video states that the formula is then used to compute sample size.

  5. Step 5: Perform the calculation in Excel

    • Provides an Excel workflow:
      • “fill in the fields for each parameter”
      • the spreadsheet then returns the sample size
    • It mentions a link and that the Excel file can be downloaded (from the video description) to calculate different sample sizes.

Speakers / Sources Featured

  • No specific individual name is given in the subtitles; the speaker is referenced only implicitly as the channel creator/instructor.
  • Source types mentioned:
    • confidence level lookup tables/statistics used to map confidence level to a z-value
    • no external author/source named

Original video