StatBioLec - PreLims

• Parts of a Box-and-whisker plot

  A) Maximum

  B) Upper quartile

  C) Interquartile range

  D) Median

  E) Lower quartile

  F) Minimum

  
• Arithmetic mean - the simple mean for continuous or discrete data
• Geometric mean - mean for growth rates, multiples
• Harmonic mean - mean for measurements in relation to units (ex. speed, velocity)
• True or False: A lower degree of freedom indicates a more precise and less variable set of data

  False
• Standard error - spread of the mean
• Normal distribution - aka z-distribution. Can be identified by the Standard deviation and Mean
• Research design - serves as a blueprint for your scientific investigation
• Observational - type of research design where individuals are observed or certain outcomes are measures. No attempt is made to affect the outcome
• Experimental - type of research design where the investigator artificially manipulates study factors or subjects
• Meta-analysis - type of research design where you combine the results from multiple studies
• Identify what type of Observational design it refers to:

  Descriptive in nature, it's the simple account of "one" phenomena wherein interesting details are described

  Case-series
• Identify what type of Observational design it refers to:

  Data is collected after the date of interest. It attempts to explain a phenomenon from a number of precious events

  Case-control
• Identify what type of Observational design it refers to:

  Analyze data collected at one time rather than over a period of time

  Cross-sectional
• Identify what type of Observational design it refers to:

  Subjects are followed over a certain period to observe a potential risk factor

  Cohort
• Identify what type of Observational design it refers to:

  In simple terms, Looking at different cases or stories one at a time over time.

  Case-series
• Identify what type of Observational design it refers to:

  In simple terms, comparing people with a problem to people without a problem to figure out what caused it.

  Case-control
• Identify what type of Observational design it refers to:

  In simple terms, taking a snapshot of a group of people to understand what's happening right now.

  Cross-sectional
• Identify what type of Observational design it refers to:

  In simple terms, following a group of people over time to see how things change for them.

  Cohort
• Identify what type of Experimental design it refers to:

  An 'experimental' is compared to another of known value (a control)

  Controlled trials
• Identify what type of Experimental design it refers to:

  Experience with an experiment is described but not compared with another treatment. It lacks a control group

  Uncontrolled studies
• Identify what type of Experimental design it refers to:

  Epitome of experimental research designs. The participants are randomly assigned to the experimental group and control group

  Randomized control trial
• Identify what type of Experimental design it refers to:

  It uses the same group as subjects and controls in 2 faces of an experiment

  Self-controlled study
• Identify what type of Experimental design it refers to:

  Experiments can use external controls. It compares a current group with a historical group that was not exposed to the new intervention

  Historical control study
• Technical replication - repeat the same measurement from the same sample or individual. It assesses the measurement error only
• Biological replication - repeat experiment from different independent examples
• Pseudo-replication - misleading practice when researchers treat single sample as multiple
• Skewness - how the sample differs in shape from a symmetrical distribution
• Kurtosis - extent to which data are distributed "tails vs. center"
• Identify

  A) Medium-tailed

  B) Uniform

  C) Leptokurtic

  D) Platykurtic

  
• Normality tests - the better, surer, way on finding out if your data is normal or not
• Shapiro-wilk test - best power when using same probability significance compared to other two tests. Asks the question, "Is the data set distribution symmetrical in nature?
• Jarque-Bera test - can give the exact interpretation for the skewness and kurtosis
• Anderson-Darling test - measures how data fits a specific type of distribution, not just the normal distribution
• p-value - the probability of getting a result as ore more extreme than observed result, assuming hypothesis is true
• Log transformation - transformation method that is usually used for data sets with high variance
• Square-root transformation - transformation method that is usually used for data sets that are highly skewed
• Log transformation - transformation method that compresses high values and spreads low values by expressing the values as orders of magnitude
• Square-root transformation - transformation method that can convert data from Poisson (discrete) distribution to a normal (continuous) distribution