Home Teachers | Mentors | Discussions | Research | Find
   The Winter Bird Feeder SurveyGuided Research Data Analysis   
The PathFinder Science Network

Creating
 the Context

  Home
  Research Focus
  Background Info
  Research Method
  Data Submission
  Results of Study
  Data Analysis
  Conclusion
  Further Research


Guided
 Research

  Research Question
  Background Info
  Research Method
  Data Submission
  Results of Study
  Data Analysis
  Conclusion
  Further Research
  Research Values


Tools
  Doing Research
  Discussions
  Email List
  Bird Links
  2004 Bird Map

   

How Can I Make Sense of My Data?

For the last several years, the results of the Kansas Winter Bird Feeder Survey have been tabulated by Dr. Elmer Finck at Emporia State University. Dr. Finck supplied this summary of five recent years to give you an idea of trends. Most notable has been the ever-increasing wintering populations of House Finches, which were not even in the top ten in 1988, when this survey began, but now are commonly amoung the top three for total numbers observed. Look at the following table to see how your own list compares:

Rank for each year 1991 - 1995

Species 1991 1992 1993 1994 1995 Mean
House Sparrow 1 1 1 1 1 1
American Goldfinch 3 2 2 2 2 2.2
Dark-eyed Junco 2 5 3 5 4 3.8
European Starling 5 3 4 4 5 4.2
Northern Cardinal 4 4 6 6 6 5.2
House Finch 10 6 5 3 3 4.2
Black-Capped Chickadee 8 7 9 7 7 7.6
Blue Jay 7 9 8 8 8 8.0
American Tree Sparrow 6 8 - 10 10 -
Harris Sparrow 9 10 10 9 - -
Pine Siskin - - 7 - - -
Mourning Dove - - - - 9 -
Number of Counts 1118 925 912 787 700

Means and Extremes

The methods and number of steps used in gathering scientific knowledge may vary from one investigator to the next, but scientific methods usually involve the alternation of two types of activities, the observational and the explanatory. So far we have been making observations about birds at our feeders, but we have not yet participated in the explanatory part of science. This experiment was set up to test a hypothesis;

H0: There is no measurable relationship between the type of seed put out and the species of birds observed at the feeder.

H1: There is a measurable relationship between the type of seed put out and the species of birds observed at the feeder.

The data set you have does not really mean very much yet because we need to compare and interrupt our data. Statistics refers to a set of procedures and rules for reducing our large data sets to manageable proportions and allowing us to take the next step and draw conclusions from our data. Our purpose is to increase the meaning of our observations reflected in our data so we will employ descriptive statistics. The most common numerical way to look at data is by means and extremes. The mean is the sum of the data values divided by the number of data points. In everyday language, this is the average. For example, the table above ranks the birds observed based on total birds. Once the rank was established over a period of years, the mean rank for each bird was calculated.

Rank your birds, with the most numerous being number one. Once you have your ranks, use the table above to see how your data compares to the past several years. Find the extremes in your data set. Look for the highest and the lowest number of a specific bird for each year.

The must effective way to interpret your data is to construct a two-variable graph for each bird species. Download as much data as is available and graph by bird species, bird rank over seed type.

Graphs are one way to visualize data and to help the researcher look for patterns. A graph is used to show the relationships of data collected from the experiment. Graphs must be constructed accurately and according to accepted rules. Usually, a graph shows the relationship between two kinds of data. These data are called variables. In this research the variables are the abundance of a bird species (rank) and the seed type. These are two types of variables.

The independent variable is data that influences the outcome of the experiment. Often the experimenter has control over this variable. Time is a very common independent variable. This data is plotted in the horizontal axis, x axis. In our research we are going to explore the relationship of seed type and species of bird. seed type is our independent variable.

The dependent variable depends on the conditions of the investigation, and frequently on the independent variable. The dependent variable is sometimes referred to as the outcome variable. The dependent data is plotted on the vertical axis, the y axis. In our research the species of bird is our dependent variable.

Remember when you make graphs;
1) Select scales for the horizontal and vertical axes which will reflect the precision of the measurements. Display the data in a proportional way. Remember, each square on the graph is equal to an assigned quantity but the scale of either axis may be changed if the graph is too compact or needs to be expanded.
2) It is important to label both the vertical and horizontal axes with the variables being graphed and also to indicated the units being used.

Spreadsheets will offer you a graphing options for your data but it is very important that you understand the graph you have made and that the graph accurately represents your data.

Construct a bar graph of the data you have downloaded.
The following is an example bar graph.


While bar graphs are interesting and a good way to visualize data, they have some problems. To learn more about graphing and how to make various kinds of graphs, the DIGSTATS site should be helpful.


Making a correlation

What is a Correlation? DIGSTATS Correlations We are interested in the realtionship between two variables on this research. Correlation is a measure of association between two variables, in this case Bird Species is one vaiable and Seed Type is the other. Correlations vary from -1 to 1, with 0 being a random relationship, 1 being a perfect positive linear relationship, and -1 being a perfect negative linear relationship. The correlation coefficient (r) also known as Pearson's r, is used to describe the strength of the relationship between the variables.

© 1996-2006 PathFinder Science