N

Data Results & Analysis

 

 

Your data and observations are presented in this section, as well as the results of any analyses performed on those data and observations. Often data are presented in the form of tables and figures. This is appropriate only if the tables and figures are referred to in the text of the Results section, which should be written as a narrative in the past tense. Each table and figure should be constructed with a descriptive and properly placed caption.Please include at least one table that shows summary statistics for your data sets and at least one figure/graph that displays results, patterns or trends in your data (e.g., an X-Y scatterplot for a correlation analysis, or a bar or column graph that compares the means of multiple samples). 

This means your tables should have the following characteristics:

• Information that is well organized into columns and rows

• Variables that are properly labeled

• Variables that contain metric units (if needed)

• A complete caption/legend that …

  • Begins with a table number (e.g., Table 1. …)

  • Appears above the table

  • Describes what the table is showing

  • Provides any additional or missing information that is needed to fully understand the table 
     

Your graphs should have the following characteristics:

• Properly labeled axes with the first letter of the first word capitalized

• Properly abbreviated metric units -- given in parentheses immediately after the label -- on the X-axis and/or Y-axis (if needed)

• Axes with proper dimensions so that the data points fill the graph

• Axes with properly spaced tick marks

• The dependent variable on the Y-axis

• A complete caption/legend that …

  • Begins with a figure number (e.g., Figure 1. …)

  • Appears below the graph

  • Describes what the graph is showing

  • Provides any additional or missing information that is needed to fully understand the graph

The first table to appear in a paper should be labeled as table 1, then next as table 2, and so on. Likewise, the first figure should be labeled as figure 1, the next as figure 2, and so on. It is strongly recommended that you look at several examples in scientific journals.

If you are using means (averages), then standard deviation and standard error (SE) should be calculated using Excel (Open Office).

• Mean – the average of all data entries

  • Add up all entries and divide by the total # of entries

  • Use the Average function in Excel, highlight the cells you want to average and hit enter

• Standard deviation (s) – the spread or dispersion in your data

  • Use the STDEV function (found in the statistical category in Excel), highlight the cells and hit enter.

• Standard Error (SE) – the standard deviation of the mean

  • SE = s/square root of n, where s = standard deviation and n = sample size

  • In the cell where you want to calculate SE, click on the cell where standard deviation was calculated, then use the divide symbol (/) type in SQRT and enter the sample size in parentheses (sample size) – it should look something like this =C14/SQRT(10); C14 was the cell that contained the already calculated standard deviation, / means divide by, SQRT means take the square root of, and (10) was the sample size OR

  • The calculation might be entered as =STDEV(C2:C11)/SQRT(10); C2 though C11 are the cells that contain the data that you want to consider for calculation

  • You may then use the SE values for the + and – on your error bars

• 95% Confidence Interval (CI) - a 95% critical value of a student t-test distribution. This is a more accurate statistic for error bars.

  • 95% CI = SE x t p(n-1)

  • Multiply the calculated SE value by the value of t at P= 0.05 (from a t-test table) for the appropriate degrees of freedom (df) for your sample (n-1)

  • If my sample size was 10, I would look at the t-test distribution table at 9 (10-1) under p = 0.05 and the value is 2.26. If my SE was calculated at 1.25 then the calculation would be =1.25*2.26

  • If 95% CI is calculated, then use these values as your + and – for the error bars.

 

 

 

 

Consult the LabWrite website for resources on graphing, calculating mean (averages), standard deviation, SE (standard error), and adding error bars to your graphs. Go to Using Error Bars in your Graph.
 

Here are some points to consider when writing your Results section:

• Text is written as a narrative in the past tense.

• Information is presented in an order or sequence that makes sense.

• Tables and figures are cited in the text.

• Tables and figures are complete with well-written captions. [Note: Figures should have properly labeled axes with units.]

• Statistical results such as the test statistic (e.g., chi square value), degrees of freedom, and probability (e.g., P = 0.071) are given.

• Writing is clear, concise, and free of grammatical errors.

• Text is free of extraneous or trivial information.

 

 

Consult the Purdue Online Writing Lab (OWL) for resources on constructing figures and graphs.