BIOZONEscimethod - 1 Steps in Biological Investigations of...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 1 Steps in Biological Investigations of your findings so far. You are now ready to begin a more indepth analysis of your results. Never under-estimate the value of plotting your data, even at a very early stage. This will help you decide on the best type of data analysis (see opposite). Skills in Biology By this stage, you will have completed many of the early stages of your investigation. Now is a good time to review what you have done and reflect on the biological significance of what you are investigating. Review the first page of this flow chart in light Photos courtesy of Pasco Observation Pilot study Something ... Lets you check ... • Changes or affects something else. • Equipment, sampling sites, sampling interval. • Is more abundant, etc. along a transect, at one site, temperature, concentration, etc. than others. • How long it takes to collect data. • Is bigger, taller, or grows more quickly. • Problems with identification or other unforeseen issues. Research Analysis To find out ... Are you looking for a ... • Basic biology and properties. • Difference. • What other biotic or abiotic factors may have an effect. • Trend or relationship. • Its place within the broader biological context. • Goodness of fit (to a theoretical outcome). Be prepared to revise your study design in the light of the results from your pilot study Variables Hypothesis Next you need to ... Must be ... • Identify the key variables likely to cause the effect. • Testable • Able to generate predictions • Identify variables to be controlled in order to give the best chance of showing the effect that you want to study. so that in the end you can say whether your data supports or allows you to reject your hypothesis. BIOZONE International © 2004 All Rights Reserved 2 Finding how one factor affects another Trend Non-Linear: The data do not plot in a straight line (i.e. curved). Example: oxygen consumption at different temperatures. Plot a scatter graph Normal data Example: Wing length vs tail length in birds. Testing for a correlation Spearman correlation coefficient Non-normal data What kind of test? Testing for a difference between groups (e.g. habitats or treatments) More than two groups of data Normal data Difference Plot a bar graph START HERE ANOVA (Analysis of variance) Paired t-test Example: Comparison of ratios of arm to leg length in chimpanzees and gorillas. Unpaired t-test Two groups of data What kind of data are you recording? Example: Frequency of occurrence of different species at two sites. Example: Survival of weevils in different pasture types. Same individuals Measurements or counts Pearson correlation coefficient Different individuals Example: Suitability of clay pots and plastic pots for plant growth. Mann-Whitney U-test Non-normal data Data must be ranked in order of increasing size. Example: Size of fruit from a plant species grown in two different habitats. Frequencies (counts only, not measurements) Comparing observed counts to an expected count Chi-squared test Example: An expected genetic ratio, or preference for different habitats. Test for goodness of fit Key to Abbreviations: CI = confidence interval Testing an association between groups of counts Chi-squared test for association Example: Association of one plant with another in an area. BIOZONE International © 2004 All Rights Reserved Skills in Biology Testing for a relationship between variables Calculate mean and 95% CI from replicates Regression Linear: The data plot in a straight line (uncommon biologically). Example: clutch size vs body size in Daphnia. ...
View Full Document

{[ snackBarMessage ]}

Ask a homework question - tutors are online