Biology research design and analysis

• The null hypothesis is the opposite of what we expect to happen.
• Developing Scientific Questions:

  • Background knowledge, observation, hypothesis, prediction, test (experiment), modify hypothesis
  • Not consistent, consistent, theory
• Data types:

  • Nominal/Categorical: classes, categories, or attributes, textual in nature (e.g. flower color)
  • Ranked/Ordinal: integers reflecting hierarchy in classification (e.g. birth order)
  • Measurement:
  • Discrete: numerical and fixed in nature, such as the number of petals on flowers
  • Continuous: assume any number of values between two points (e.g. height, length, mass, etc.)
• Observational Studies:

  • Compare variables measured from different conditions, areas, etc.
  • Tend to be the first data collected
  • Helps generate hypotheses
  • Rarely address cause/effect
  • Also used to monitor or evaluate status
  • Foundation of knowledge
• Comparative Studies:

  • Independent variable varies naturally within a system of interest
  • Allows to formally test hypotheses
  • Limitations:
  • Small sample size
  • Confounding variables
  • More typically used in ecological or physiological studies
• Perturbation/Response Studies:

  • Utilizes natural conditions following large-scale disturbances
  • Natural disasters, human-caused disturbances
  • Almost best considered a special type of descriptive study
  • Similar limitations
• Manipulative Experiments:

  • Typically the most familiar type of study
  • Impose treatment or treatments, then observe response to the treatment(s)
  • Independent (predictor) variable, dependent (response) variable
  • Limitations:
  • Are reductionist
  • Small numbers & short-time scales
  • Ethical constraints
• Deductive Science/Modeling:

  • Specify values (parameters) for variables or conditions
  • Use logic and math to predict outcome
  • Parameters often derived following empirical studies
  • Can compare models to experimentally collected data
  • Helpful to validate the model
  • Identify gaps in knowledge
• Presenting Data:

  • Never report raw data
  • Present summaries of the data (descriptive statistics)
  • Tables and Figures:
  • Table: arrangement of data into rows and columns
  • Figure: any other type of graphical representation (e.g. graphs, photos, maps, etc.)
• Sources of Variation:

  • Random Error Variation
  • Treatment Effects
  • Experimental Artifacts
• Minimizing Experimental Artifacts and Error:

  • High degree of precision & accuracy
  • Effective controls
  • Absence of bias
• Controls:

  • Do not receive the treatments
  • By comparing the treatment to control, variation cancels out and differences are due to treatment effect
  • Not all designs may have a "control" group
• Sample Size = Level of Replication:

  • A replicate is a repeated unit
  • Replicates allow us to control & quantify random variation, estimate population parameters
  • More replicates mean more reliable estimates
• Randomization:

  • Assigning individuals to different treatment groups randomly
  • Eliminates systemic sources of bias
  • Ensures independence of data
• Statistics:

  • Descriptive statistics: summary statistics, organization and summarization of data displayed as tables and figures
  • Inferential statistics: drawing conclusions beyond what is seen in the data alone
• Writing by Biologists:

  • Research proposals, scientific papers, grant & paper reviews
  • Teaching: lectures, exams
  • Misc.: emails, memos, correspondence
• Literature Cited:

  • Use author, year format for in-text citations
  • Be concise in citing references
  • Cite only sources read and confident discussing
  • Avoid citation overkill
• Back to... Statistics:

  • Descriptive statistics: summary statistics, organization and summarization of data displayed as tables and figures
• What we are trying to do:

  • Determine if a predictor variable(s) has an effect on the response variable
  • Solid experimental design approaches are necessary
  • Methods and experiments allow sampling populations for analysis
  • Purpose is to make conclusions about a group of measurements of a variable being studied
• Difference between a population and a sample in statistics:

  • Sample: group of individuals randomly selected from a larger group, described by statistics
  • Population: all organisms comprising the group of interest, described by parameters
• Statistics give us a common language
• Statistics allow us to test hypotheses
• We use statistics to determine if the effects we see are real or not
• Statistics provide information about data to help understand findings (descriptive statistics)
• Statistics help draw conclusions beyond what is seen in the data alone (inferential statistics)
• Statistics help determine what a sample tells us about the population
• Statistics help determine if a treatment made a difference
• Writing by Biologists includes:

  • Research proposals for research funds
  • Scientific papers
  • Grant & paper reviews
  • Teaching through lectures, handouts, and exams
  • Miscellaneous writing like emails, memos, correspondence, and outreach
• Good writing skills are crucial for scientists
• Writing basics:

  • Understand your topic
  • Have a writing plan
  • Write to illuminate, not to impress
  • Write for your audience
  • Make a statement and then back it up
  • Distinguish between fact and possibility
  • Don't plagiarize
  • Revise your work
• Revising Writing Basics:

  • Stick to the point
  • Say exactly what you mean
  • Never make the reader back up
  • Don't make readers work too hard
  • Be concise
  • Proofread
• Polishing Writing Basics:

  • Always underline or italicize species names
  • Remember "data" is plural
  • Appearances matter
  • Keep a copy of your final product
  • Never let style & technology become more important than the content
• Literature Cited:

  • Use author, year format for in-text citations
  • Be concise in citing references
  • Cite only sources you have read and can discuss confidently
• Descriptive statistics involve:

  • Summary statistics
  • Organizing and summarizing data
  • Displaying data as tables and figures
• Statistics help determine if predictor variables affect response variables
• Difference between population and sample in statistics:

  • Sample: randomly selected group of individuals from a larger group, described by statistics
  • Population: all organisms in the group of interest, described by parameters
• Statistics help determine if two samples come from different populations or the same population
• Descriptive Statistics:

  • Frequency distributions show the occurrence frequency of variable values
  • Histograms display data distribution
• Characterizing populations involves:

  • Central Tendency (Mean, Median, Mode)
  • Dispersion (Variance, Standard Deviation)
• Symbols in statistics:

  • Mean of a population = μ, Mean of a sample = x̄
  • Standard deviation of a population = σ, Standard deviation of a sample = s
  • Variance of a population = σ^2, Variance of a sample = s^2