My data was represented graphically. I calculated the mean of the total number of Allium cernuum samples obtained from the three sites per quadrats, computed the mean and represented it graphically. I think it was best to represent the data in this manner because it will permit a statistical test to be performed which in this case was an ordinary one-way ANOVA test due to the fact that the analysis was being performed across three different data sets. This test will aid in showing the significant differences in terms of the total abundance of the plant relative to the site farthest from the industrial site. I also chose to represent the data in tabular form. In this case, I calculated the density and frequencies of the Allium cernuum plant in each site and represented the values in the table. The density gives a measure of the species abundance while the frequency denotes how often the plant is frequently recorded in a quadrat.
Category: Post 8: Tables and Graphs
Blog Post 8: Tables and Graphs: Cates Park
Although I intend to focus on the success of Tsuja heterophylla on nurse logs versus the forest floor, I collected data on the presence and absence of all species found within the quadrats I studied in four regions of Cates Park in North Vancouver. Limiting the data to one species in a chart helps with the ease of interpretation, but limits the understanding of species richness in the microsuccessions of nurse logs versus the forest floor. Organization was simple due to the presence and absence data collected, and would have been more difficult had I included other species present. I will likely discuss this in my final paper to address the succession species found in Cates Park. The outcome was as I expected: Tsuja heterophylla were present more often on nurse logs than not. Further exploration could include canopy cover in relation to this tree’s success or competition with other species in the region.
Blog Post # 8 – Tables and Graphs
Create a blog post discussing your table or graph. Did you have any difficulties organizing, aggregating or summarizing your data? Was the outcome as you expected? Did your data reveal anything unexpected or give you any ideas for further exploration?
Finally finished my data collection and got it all organized into an excel spreadsheet! I had to re-evaluate my target of 150 bird observations. Between working full time, balancing other commitment, and good ol’ Alberta weather systems, finding opportunities to get to my observation site within the time criteria I had pre-defined proved more challenging than anticipated! I did manage to get 90 birds though, (30 for each time of day) so I feel like I still have a good set of data to work with.
I started playing with my data this weekend, sorting it by species, gender, time of day, occurrence along my gradient (land, shallow water, deep water). I tried to emulate graphs that I saw in some of the studies I’ve been reading as part of my literature review, as well as playing with various functions available on the excel graphing software. Luckily I’m pretty comfortable working with excel, so aggregating data wasn’t much of an issue. I think my main challenge will be focusing on exactly WHAT aspect of the data I will use in my paper.
Is there such thing as too much data?
I do find myself going off on tangents…lets see what happens when I add up this data, or graph these groups together. I come from a medical background (Clinical Pharmacist) and was always taught that bad statistics can be used and abused to say whatever you want (many a large pharmaceutical company has published questionable trial results that are in their favor!) and I can see how this could occur as I think up different things to graph, and look for patterns that pop out.
My plan is still to stick to comparing my birds’ behavior patterns to that of dabbling ducks published in existing literature. So far nothing has really jumped out at me as “unexpected” but I’m still compiling my literature so we’ll see what happens!
An example of some of my graphs so far:
` 
Blog Post 8: Table and Graphs
The following graph was created based on the data collected for the field research project:
This graph shows the total number of bird species (bar chart) observed with the ambient air temperature overlaid (line chart). As there were only two variables collected for this research project it was fairly easy to layout the data in a graph.
The outcome was not as I was expecting. I was expecting an linear increase in the number of bird species observed as temperatures increased. It would be interesting to see if the same trend exists if more data is collected.
Blog Post 8: Tables and Graphs
I chose to create a graph to represent the data I collected. It was beneficial that we had created a graph/table for our hypothesis as I learned from that what information I needed to include and I also learned some important lessons about the type of graph to use based on my study design. I had done my hypothesis graph incorrectly so I needed to do some research to figure out the type of graph that I should use. My response variable (birds) is a categorical variable and my predictor variable (temperature) is a continuous variable. Therefore my study design is a logistic regression. Initially I had created a line chart. However, a line chart would suggest that the response variable was continuous. I did some research and in the end I decided that a point graph (called a scatter plot in some journals) was the best way to chart my observations and results. I also had trouble with the legend and knowing what information should be included there. I tried to create it so that if an individual read the chart without knowing any information about my research study they would be able to decipher the results. I think a table would have given more information than I was able to include in the graph such as the times of day that the observations were collected or information on other variables such as cloud cover and wind. However, I chose to create a graph because the visual information included in a figure can often tell a story that words cannot. My results surprised me and suggest that my hypothesis is not true, or rather that the null hypothesis may be correct. I would be interested in doing further research on the subject of bird activity with weather and expand it beyond temperature alone. How much of an affect is cloud cover on bird activity or humidity? Also, do smaller birds have different tolerances to weather than larger birds? Lastly, is bird activity affected by a combination of variables that include temperature but also precipitation, wind, cloud cover, humidity and time of day?
Blog Post 8: Tables and Graphs
I did not have too much troubles organizing my graphs. The outcome was as expected for small birds but the data did not seem to signify anything significant for large or medium sized birds. This may be because weather effects small birds more so than larger birds, but I have not found anything that confirms this and will have to continue my research to see if this is a potential explanation.
Blog Post 8: Tables and Graphs
I used regression analysis to quantify the correlation between birch distribution over varying aspects across a hillslope. On the graph I included a logarithmic trendline and the coefficient of determination (R squared) value. A R^2 value of 0.4198 does not indicate a strong correlation between aspect and birch distribution.
I didn’t have any trouble organizing my data and felt like this was a straightforward exercise. When I first started this project I expected to see a stronger correlation between aspect and birch distribution but now I think there are other factors that are more influential than aspect alone. I still think that soil moisture is a very dominant factor and is correlated with aspect, however; drainages, depressional terrain, or other areas where water accumulates will likely create suitable habitat for birch trees regardless of aspect.
That being said, this is a small sample size over a small area. More samples over a greater area may yield a stronger correlation between aspect and the distribution of birch.
Post 8: Tables and Graphs
I used an ANOVA analysis using excel to plot plot distance from a path (categorical variable) to mean ratio of trees to shrubs (continuous variable), inclusive of standard deviation, and found statistical significance. I was significantly stressed about doing this because I haven’t yet taken a stats class but I found some resources (thank you Percy) and learned quite a bit about statistics that will help me in the classI have to come.
With my hypothesis I am testing for a number of variables and decided to present this relationship because it is statistically significant. However, most of the tests did not disprove the null-hypothessis. I was wondering what I did see during initial observation. I still think there may be a relationship. Many of my p-values were low but not near the 5% threshold, lending themselves to show more of a relationship than not. I am doubtful of the 5% p-value rule and have read academics challenging the 5% threshold. I have much more to learn.
Post 8: Tables and Graphs
I had no difficulties organizing my data. However, I had difficulties in finding the best way to summarize and illustrate the data. It took a little playing around with the data to find the best outcome which summarizes and explains what the overall study was about. The outcome was what I expected, there is a significant difference between area with <50% canopy cover, compared to the other levels of canopy coverage. Each dot represents the sample means of each strata (canopy cover). each interval is a 95% confidence interval that the group mean is within the groups confident interval. The error bars represent the standard error of the means, in other words, how accurate the sample represents the population. My study focuses on the moss abundance on the ground, I wonder if this pattern would be the same if I looked at moss abundance on trees. If the pattern was the same, I wonder if the same elements play the same roles. On the other hand, if it was different, would it be due to the same mechanisms driving this pattern, or are other factors included.
Blog Post 8 – Tables and Graphs
The data that was collected was separated into 4 different categories. The first three are the locations where the species observations were made (House of Learning, Science Building, House 9), and the fourth was a table of how many people were seen in each location over a span of an hour. For all four tables, the averages (means) were calculated and bar charts were used to summarize the data.
The tables generally supported the hypothesis however, further statistical analysis will help determine this.
