Scribd Logo
Upload

Welcome to Scribd!

  • 58 views

    Lecture 15

    Uploaded by

    m_michael_c
    This document discusses sample size estimation for descriptive studies. It defines descriptive studies as aiming to estimate parameters without testing hypotheses. The key steps are: 1) identify variables of interest, 2) determine the type of estimates (e.g. means, proportions), 3) select populations/subgroups, 4) indicate expected population values and estimate standard deviations, 5) choose a confidence level, 6) define a precision level, and 7) compute the sample size based on these assumptions. Formulas and examples are provided for estimating sample sizes for one group proportions and means, and for comparing two groups' proportions and means. An Excel spreadsheet can assist with calculations.

    Copyright:

    © All Rights Reserved
    Download as PPT, PDF, TXT or read online from Scribd

    Lecture 15

    Uploaded by

    m_michael_c
    58 views21 pages
    This document discusses sample size estimation for descriptive studies. It defines descriptive studies as aiming to estimate parameters without testing hypotheses. The key steps are: 1) identify variables of interest, 2) determine the type of estimates (e.g. means, proportions), 3) select populations/subgroups, 4) indicate expected population values and estimate standard deviations, 5) choose a confidence level, 6) define a precision level, and 7) compute the sample size based on these assumptions. Formulas and examples are provided for estimating sample sizes for one group proportions and means, and for comparing two groups' proportions and means. An Excel spreadsheet can assist with calculations.

    Original Description:

    15

    Copyright

    © © All Rights Reserved

    Available Formats

    PPT, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document discusses sample size estimation for descriptive studies. It defines descriptive studies as aiming to estimate parameters without testing hypotheses. The key steps are: 1) identify variables of interest, 2) determine the type of estimates (e.g. means, proportions), 3) select populations/subgroups, 4) indicate expected population values and estimate standard deviations, 5) choose a confidence level, 6) define a precision level, and 7) compute the sample size based on these assumptions. Formulas and examples are provided for estimating sample sizes for one group proportions and means, and for comparing two groups' proportions and means. An Excel spreadsheet can assist with calculations.

    Copyright:

    © All Rights Reserved
    Download as PPT, PDF, TXT or read online from Scribd
    Download as ppt, pdf, or txt
    58 views21 pages

    Lecture 15

    Uploaded by

    m_michael_c
    This document discusses sample size estimation for descriptive studies. It defines descriptive studies as aiming to estimate parameters without testing hypotheses. The key steps are: 1) identify variables of interest, 2) determine the type of estimates (e.g. means, proportions), 3) select populations/subgroups, 4) indicate expected population values and estimate standard deviations, 5) choose a confidence level, 6) define a precision level, and 7) compute the sample size based on these assumptions. Formulas and examples are provided for estimating sample sizes for one group proportions and means, and for comparing two groups' proportions and means. An Excel spreadsheet can assist with calculations.

    Copyright:

    © All Rights Reserved
    Download as PPT, PDF, TXT or read online from Scribd
    Download as ppt, pdf, or txt
    of 21

    SAMPLE DESIGN: HOW

    MANY WILL BE IN THE


    SAMPLEDESCRIPTIVE
    STUDIES ?
    Lu Ann Aday, Ph.D.
    The University of Texas
    School of Public Health
    TYPES OF OBJECTIVES
    DESCRIPTIVE ANALYTICAL
    Describes Explains
    Is more exploratory Is more explanatory
    Profiles characteristics Analyzes why group(s)
    of group(s) have characteristics
    Focuses on why?
    Focuses on what?
    Assumes an hypothesis
    Assumes no hypothesis
    Requires comparisons
    Does not require (between groups or
    comparisons (between over time)
    groups or over time) Try to maximize
    Try to maximize power to detect
    precision of estimates differences, if they exist
    RELATING SAMPLE SIZE ESTIMATION
    TO STUDY OBJECTIVES
    Select the sample size estimation procedure that
    best matches the study design underlying the
    respective study objectives
    Compute the sample size required to address each
    objective
    Based on the sample sizes required to address each
    of the objectives, appropriate sample size
    adjustments, as well as time and resource
    constraints, recommend an overall sample size
    Discuss possible limitations in terms of statistical
    precision or power in addressing any specific study
    objective(s), given the recommended sample size
    CRITERIA:
    Descriptive Studies
    Objective:
    to estimate a parameter, i.e.,
    provide a precise estimate for
    selected variable(s)
    Framework:
    normal sampling distribution
    NORMAL SAMPLING
    DISTRIBUTION
    Sampling Distribution: distribution of
    estimates, e.g., mean, for all possible
    simple random samples of a certain size
    that could be hypothetically drawn from
    the target population
    Population Mean: grand mean of all
    possible simple random samples of a
    certain size that could be hypothetically
    drawn from the target population
    STANDARD ERROR
    Definition: average variation of all
    possible simple random samples of a
    certain size that could be hypothetically
    drawn from the target population
    Formula:
    SE = s/n, where,
    SE = standard error
    s = sample standard deviation
    n = sample size
    = square root (sqrt)
    CONFIDENCE INTERVAL
    Definition: range of values in which the
    population mean is likely to be contained,
    with a given level of probability, defined
    by the standard errors of the sampling
    distribution

    Confidence Interval Standard Errors (Z)


    68 % 1.00
    90 % 1.645
    95 % 1.96
    99 % 2.58
    EXAMPLE: Mean Estimate
    Population Mean: 5 visits
    Standard Error: .50 visits
    Confidence Interval:
    68 % = 1.00 * .50 = +/- .50 visits
    90 % = 1.645 * .50 = +/- .82 visits
    95 % = 1.96 * .50 = +/- .98 visits
    99 % = 2.58 * .50 = +/- 1.29 visits
    EXAMPLE: Proportion Estimate
    Population Proportion: .50, i.e., 50%
    Standard Error: .025, i.e., 2.5%
    Confidence Interval:
    68 % = 1.00 * 2.5 = +/- 2.5%
    90 % = 1.645 * 2.5 = +/- 4.1%
    95 % = 1.96 * 2.5 = +/- 4.9%
    99 % = 2.58 * 2.5 = +/- 6.4%
    SAMPLE SIZE ESTIMATION: Cross-
    Sectional (One Group)Proportion
    Formula:
    n = Z21-/2 P(1-P)/d2, where,
    n = sample size
    Z21-/2 = confidence interval
    P = estimated proportion
    d = desired precision
    SAMPLE SIZE ESTIMATION: Cross-
    Sectional (One Group)Proportion
    Example:
    n = Z21-/2 P (1-P)/d2
    n = 1.962 * .50(1-.50)/.052
    n = 384

    Note: See Table 7.1B, Aday & Cornelius,


    2006, for sample size estimates based on
    different estimated proportions (P) and levels
    of desired precision (d).
    SAMPLE SIZE ESTIMATION: Cross-
    Sectional (One Group)Mean
    Formula:
    n = Z21-/2 2/d2, where,
    n = sample size
    Z21-/2 = confidence interval
    = estimated standard deviation
    d = desired precision
    SAMPLE SIZE ESTIMATION: Cross-
    Sectional (One Group)Mean
    Example:
    n = Z21-/2 2/d2
    n = 1.962 * (2.5 2) /1 2
    n = 24
    Note: To estimate when not known, estimate
    the inter-quartile range by dividing the possible
    range of values by 4, e.g., if range is 0-10, then
    10/4 = 2.5.
    SAMPLE SIZE ESTIMATION: Group-
    Comparison (Two Groups)Proportion
    Formula:
    n = Z21-/2[P1(1-P1) + P2(1-P2)]/d2, where,
    n = sample size
    Z21-/2 = confidence interval
    P1 = estimated proportion (larger)
    P2 = estimated proportion (smaller)
    d = desired precision
    SAMPLE SIZE ESTIMATION: Group-
    Comparison (Two Groups)Proportion
    Example:
    n = Z21-/2 [P1(1-P1) + P2(1-P2)]/d2

    n = 1.962 [(.70)(.30) + (.50)(.50)]/.052

    n = 707 (in each group)


    SAMPLE SIZE ESTIMATION: Group
    Comparison (Two Groups)Mean
    Formula:
    n = Z21-/2 [22]/d2, where,
    n = sample size
    Z21-/2 = confidence interval
    = estimated standard deviation
    d = desired precision
    SAMPLE SIZE ESTIMATION: Group
    Comparison (Two Groups)Mean
    Example:
    n = Z21-/2 [22]/d2
    n = 1.962 * [2 * (2.5 2)] /1 2
    n = 48 (in each group)
    Note: To estimate when not known, estimate
    the inter-quartile range by dividing the possible
    range of values by 4, e.g., if range is 0-10, then
    10/4 = 2.5.
    SUMMARY: Steps in Estimating
    Sample Size Descriptive Studies
    1. Identify the major study variables.
    2. Determine the types of estimates of
    study variables, such as means or
    proportions.
    3. Select the population or subgroups
    of interest (based on study objectives
    and design).
    4a. Indicate what you expect the
    population value to be.
    4b. Estimate the standard deviation of
    the estimate.
    SUMMARY: Steps in Estimating
    Sample Size Descriptive Studies
    5. Decide on a desired level of
    confidence in the estimate
    (confidence interval).
    6. Decide on a tolerable range of
    error in the estimate (desired
    precision).
    7. Compute sample size, based on
    study assumptions.
    SAMPLE SIZE ESTIMATION:
    EXCEL SPREADSHEET
    See EXCEL file with spreadsheet
    for computing sample sizes.

    You might also like