Research methods

• Scientific research is a fundamental part of Environmental Science and good research skills are needed to collect representative data so that valid conclusions can be reached
• Good decisions need good quality information to support the decision-making process
• Attributes of quality of methods and data

  • Quality
  • Accuracy
  • Precision
  • Representative sample
  • Reliable method
  • Anomalous result
  • Valid study
• Accuracy


  A measure of how close the recorded result is to the real value
• Precision


  The intervals between possible recorded results, for example, a ruler that measures in millimetres is more precise than one that measures in centimetres
• Representative sample


  A subsample that accurately reflects the complete data set
• Reliable method


  Provides consistent, accurate results
• Anomalous result


  A result that differs from other results with which it was expected to be similar
• Valid study


  Repeating the study should show whether is a real result or was caused by other variables
• Produces precise accurate, reliable results upon which conclusions can be based
• Factors to consider when designing and carrying out environmental science fieldwork, investigative enquiries and lab work

  • Appropriate risk management
  • The practical equipment and materials required and how to use these safely and correctly
  • The data to be collected and analysed
  • Using appropriate apparatus/instruments to record quantitative measurements (for example, temperature, length and pH)
  • Using appropriate apparatus/instruments and methodologies to measure abiotic and biotic factors (for example, light intensity, humidity, population size)
  • Selecting the right sampling techniques (for example, pitfall traps, Tullgren funnel, soil texture analysis, water turbidity, light traps)
  • Selecting the right statistical analysis techniques
• This chapter provides a summary of methodologies, skills, research methods, and statistical analysis techniques used in environmental investigations
• Scientific methodologies


  • Identifying a topic of interest
  • Finding out what has already been learnt
  • Formulating a hypothesis/null hypothesis that can be tested
  • Designing and carrying out an experiment to test the hypothesis
  • Designing the investigation, for example, including consideration of location of sampling sites, number of samples, size of samples, standardised technique, timing of sampling
  • Statistical analysis to assess statistical significance
  • Analysing the results and drawing conclusions
  • Planning further research to fill gaps in existing knowledge or develop the study further
• Careful planning is needed to ensure that the data collected can be used to conclude whether a hypothesis should be accepted or rejected
• All studies should be planned and carried out so that the statistical significance of the results can be assessed
• It is rarely possible to collect all the data that exists about a particular issue. It is more usual to collect sub-samples and use these to estimate the full situation. This will only produce reliable conclusions if the sub-samples are representative
• The planning of sub-sampling must consider factors that may affect how representative they are: location, timing, size, number
• A preliminary study should be used to test the reliability of the method
• Sampling methods


  • Random sampling
  • Systematic sampling
• Random sampling


  If the study area has a regular shape then a grid can be laid out and the coordinates of sampling sites selected using random numbers. If the study area has an irregular shape then all the possible sampling sites can be numbered and selected using a random number generator.
• Systematic sampling


  This involves samples taken using a chosen interval distance or spacing. The choice of sampling sites is not based on choices made using observable differences in the study area.
• The distance between sampling sites should be selected to detect variable data without collecting more data than necessary. This can be determined with a preliminary study.
• If the factor that is being measured changes with time, then it will be necessary to sample on different occasions to produce a mean result that represents the 'typical' value or to identify trends in changing values.
• The ideal time interval between samples depends upon the rate at which values change. This can be determined with a preliminary study.
• If the variable being measured is not homogenous then larger samples are more likely to produce representative results. The sample size needed can be determined with a preliminary study.
• A single sample may not be representative as there may be variability between samples. Collecting multiple samples will help to eliminate the effect of variability. The number of samples needed to produce a representative mean depends upon the degree of scatter around the mean. This can be found using a preliminary study.
• The degree of scatter of values around the mean can be assessed by calculating the Standard Deviation.
• The number of samples collected will also depend upon how the results will be analysed. It is easier to assess the degree of statistical significance of the results with larger numbers of samples.
• To allow comparison of the results that are collected at different times, in different locations or by different researchers, they must all be collected in exactly the same way.
• Other variables that may affect the results must either be controlled or monitored to assess their effects.
• Reasons for ecological monitoring

  • To find out which species are present
  • To monitor the habitat conditions, how they change, and which conditions individual species require
  • To monitor features of populations and how they change over time such as population size, age structure, growth rates, breeding rates, territory size, population movements
• Reliable, representative data are important in making good conservation management decisions.
• Ecological monitoring techniques


  • Quadrats
  • Pond net
  • Kick sampling
  • Surber samplers
  • Plankton net
  • Sweep nets
  • Aerial insect nets
  • Colonisation media
  • Pitfall traps
  • Beating trays
• Quadrats


  A study area, usually square, within which populations can be studied. The results from a number of quadrats can be scaled up to estimate the overall situation for the entire area.
• Types of quadrats


  • Open frame quadrats
  • Grid quadrats
  • Point quadrats
• Limitations of quadrats include subjective judgement may be involved, for example, estimating percentage cover, and the quadrat frame may flatten or move plants.
• Pond net


  Can be swept through the water or aquatic vegetation to capture invertebrates, amphibians, or fish. Using a particular length of sweep, or number of sweeps, can be used in an attempt to standardise the method and produce comparative data.
• Limitations of pond nets include mobile species may escape.
• Kick sampling


  A net is held with its straight edge on the river bed and its face open to the current. The riverbed in front of the net is disturbed by repeatedly scraping a boot over it. Invertebrates that are dislodged are caught in the net. Controlling the number and length of kicks can be used to standardise the method.
• Limitations of kick sampling include accurate standardisation is difficult, buried organisms or those fixed on stones may not be dislodged, some organisms can swim and avoid the net, the water flow rate is variable and affects the capture rate, and fragile organisms may be destroyed.