Biostatistics Lecture - Chapter 9

• The concept of Sampling:

  • Selecting cases
  • Making estimates
  • Relative to effort and resources
  • Estimate/predict outcome
  • Represent population
• Tradeoff - between time and resources and compromising level of accuracy.
• Characteristics of good samples:

  • Representative
  • Accessible
  • Low cost
  • Appropriate size
• Study Population (N) - group of subjects from which you select your sample.
• Sample - subset of population elements from data is collected to estimate the outcome.
• Sample size (n) - number of subjects from whom you obtain the required information.
• Accessible Population - conform to criteria and accessible for study.
• Target Population - cases which would like to generalize.
• Eligibility/Inclusion Criteria - a criteria specifying population characteristics which is done to enhance the study's construct validity.
• Exclusion Criteria - a criteria that specifies the characteristics that people must not possess.
• Sampling Design/Strategy - the way subjects are selected.
• Sampling Frame - a list of each subject in the study population from which a sample is drawn; requires all elements in a sampling population to be individually identified.
• Hypothesis - a claim or a statement about a property of a population.
• Statistic - number calculated from your data.
• Parameters - estimates arrived at from sample statistics.
• Parameters aim to find answers to research questions in the study population, not in the sample collected.
• The greater the sample size, the more representative it is of the population.
• Quantitative Sampling - samples selected to achieve statistical conclusion validity.
• Probability Sampling - each element in the population has an equal and independent chance of selection.
• Non-probobability sampling - does not allow the theory of probability. It is used when there is no sample frame and due to practicality.
• Factors affecting sample size:

  • Effect size
  • Homogeneity
  • Cooperation and Attrition
  • Subgroup analysis
• Steps in conducting a statistical experiment:

  • Formulate the question
  • Gather the data
  • Organize and analyze the data
  • Interpret
• Primary Source Data - collected by the statistician/researcher.
• Secondary Source Data - data that are already available.
• Precautions in collecting data:

  • Primary: proper collection schemes should be followed
  • Secondary: data should be organized, evaluated, and interpreted.
• Census - complete enumeration of an entire population.
• Survey - information solicited from people.
• Observation - using senses to examine people in natural settings or naturally occurring situations.
• Experiment - compare groups.
• Simulations - use of models replicating real life conditions.
• Review of documents and records - collected from existing records.
• Six Qualities of Statistical Data:

  • Timeliness
  • Validity
  • Reliability
  • Completeness
  • Precision
  • Integrity
• Data coding - transforming data into codes.
• Data encoding - the process of converting data into a form that can be stored in a computer.
• Data editing - checking for errors.
• Analyzing Data:

  • Use of statistical tools on data encoded.
  • Factors dictating choice of statistical data.
• Presenting Data:

  • Representing organized, summarized, analyzed data in the form of tables and graphs.
  • Analyze trends, compare and contrast relationships of variables.
• Master table - a table that contains the primary key of all other tables in the database.
• Dummy table - allows researchers to preview expected research results.
• Frequency distribution table - shows the actual number of distributions.