Lab 2 Data Presentation PDF

Preview

Summary

Lab manual...

Description

Spring 2019

Lab 2 Data Presentation & Interpretation Learning goals  Identify the key components of a graph and a table, and where the independent, dependent and categorical variables belong.  Determine the appropriate type and number of graphs for a given data set.  Appropriately label the x- and y-axes of a graph.  Determine how to organize a summary table.  Interpret the results or trends indicated by the data set, including error bars.  Develop a figure caption and table title that summarizes the experimental variables, comparator group/condition, methodology, sample being tested, sample size, and statistical information, as appropriate.  Identify which means are significantly different from each other based on symbols used to indicate statistical significance in graphs.  Develop an experimental conclusion summarizing the: o experimental manipulation, o comparator group/condition, if appropriate o general physiological mechanism or phenomenon (how the physiology is changing), o population to which to the results apply, o trend, and whether the results were statistically significant.  Identify potential experimental limitations, both methodological and physiological and be able to explain how they could have influenced the outcome of the experiment.

Graphs (adapted from Silverthorn 7th edition) Graphs are pictorial representations of the relationship between two (or more) variables, plotted on a rectangular region. We use graphs to present a large amount of numerical data in a small space, to emphasize comparisons between variables, or to show trends over time. A viewer can extract information much more rapidly from a graph than from a table of numbers or from a written description. A wellconstructed graph should contain (in very abbreviated form) everything the reader needs to know about the data, including the purpose of the experiment, how it was conducted, and the results. All scientific graphs have common features (Figure 1). The independent variable (variable manipulated by the experimenter) is graphed on the horizontal x-axis*. The dependent variable (variable measured by the experimenter) is plotted on the vertical y-axis. For this class, create a single graph per dependent variable. If the experimental design is valid and the hypothesis is supported, changes in the independent variable (x-axis) will cause changes in the dependent variable (y-axis) in an experimental study. Each axis of the graph is divided into units represented by evenly spaced tick marks on the axis. A label indicates what variable the axis represents (time, temperature, amount of food consumed) and in what units it is marked (days, degrees Celsius, grams per day). The origin is the intersection of the two axes. The origin usually, but not always, has a value of zero for both axes. If multiple groups are shown on one graph, the lines or bars representing the groups may have labels, or a key may show what group each symbol or color represents. Updated 9/2018

Lab 2 – Data Presentation  1

Spring 2019

Figure caption goes below the figure and includes: description of experimental variables, comparator, methodology, sample being tested, sample size, statistical test used, explanation of any symbols on graph, and level of significance.

Figure 1. Key components of a graph

A figure caption (also called a legend) is located below each graph. The caption begins with a numerical label (Figure #) and includes the following:  Description of the experimental variables (independent variable, dependent variable), including any comparator group/condition  Methodology (equipment used, how each dependent variable was measured)  Sample being tested and sample size (number of subjects)  Indication of the statistical test used  Explanation of any symbols on the graph used to indicate statistical significance and the chosen level of significance All graphs are numbered in the order they appear within the text. As a rule of thumb, the figure caption should provide enough detail so that another reader can recreate the experiment without reading any of the supporting text. If necessary, a figure caption can contain multiple sentences. A figure caption that states only “Variable X versus variable Y” or “Results of Experiment 1,” for example, is not sufficient.

Types of graphs Most graphs you will encounter in physiology display data either as bars (bar graphs), as lines (line graphs), or as dots (scatterplots). The type of graph used to present the data is determined by the independent (or grouping) variable or whether the data is paired. Bar graphs are used when the independent variables of the experiment are categorical. The bars are lined up side by side along one axis so that they can be easily compared with one another. Scientific bar graphs are traditionally 2-dimensional and have the bars running vertically. *Note: In this course, we will plot data from pre/post study designs with bar graphs, with the categories ‘before’ and ‘after’, or ‘pre’ and ‘post’, condition/treatment/intervention. Line graphs are appropriate when the independent variable on the x-axis is continuous. Each point on the graph may represent the average of a set of observations. Symbols are used to indicate the actual points at Updated 9/2018

Lab 2 – Data Presentation  2

Spring 2019

which data were collected. Because the independent variable is a continuous function, the individual data points on the graph can be connected with a line. Connecting the points allows the reader to interpolate, or estimate the values between the measured values. Extending the graph beyond the plotted points is called extrapolation. Scatterplots show the relationship between two variables, such as time spent studying for an exam and performance on that exam, and would be used with some types of descriptive studies. Usually each point on the plot represents observations of one member of a test population. A line never connects individual points on a scatter plot, but rather a “best-fit” line or curve (sometimes called a trendline or regression) may indicate a trend in the data.

Tabulating raw data When collecting data during an experiment, it is important to record it in a tabular format in an Excel spreadsheet or lab notebook. Your raw data table should include all the data you collected during your experiment as well as all necessary calculations. When organizing a raw data table, the subject number is typically placed in the first column followed by the independent variable(s) and dependent variable(s) in subsequent columns. Headers are located at the top of each column to help identify the variables. An example raw data table is shown in Figure 2. Figure 2. Image of an example raw data table

Summary tables Summary tables are often used for reporting extensive numerical data in an organized manner. The standard convention for organizing summary tables is to place the independent variable in the first column and the dependent variable(s) with standard deviations in the subsequent columns. Organizing the table in this way enables the reader to focus their effort on comprehending the data presented, rather than having to take the time to first determine the organization of the table.

The key components of a summary table are shown in Figure 3. Each column should have a heading describing the material below it. For each dependent variable, the measurement of error is indicated in the same column separated by the symbol “±.”

Updated 9/2018

Lab 2 – Data Presentation  3

Spring 2019

Table title goes above the table and includes: description of experimental variables, comparator, methodology, samples being tested, sample size, data represent average ± SD, statistical test used, and alpha level. Symbols used to indicate statistical significance.

Figure 3. Image of key components of a summary table

Summary tables typically have titles located above the data. Similar to a figure caption, the table title begins with a numerical label (Table #) followed by:  Description of the experimental variables (independent, dependent), including units, and comparator group/condition, if appropriate  Methodology (equipment used, how measured)  Sample being tested and sample size from each group (number of subjects)  Data represent average ± SD  Type of statistical test(s) used and alpha level  Explain symbols used to denote statistical significance All summary tables are numbered in the order they appear within the text.

Experimental conclusion The conclusion is a statement of what researchers think they have learned about the physiology based on performing the experiment – i.e., it is the researchers’ answer to the research question based on their interpretation of the results. In the conclusion, researchers attempt to generalize what they have learned about the effects of the experimental manipulation on the physiology, to the population being studied. Thus, the conclusion should include the:  experimental manipulation,  how the dependent variable changed in response to the manipulation,  population,  trend, and  whether the results were statistically significant. A typical format for the conclusion might read “These data suggest that {experimental manipulation – specific manipulation IV}, as compared to {comparator}, DID/DID NOT lead to a significant DECREASE/INCREASE in {DV1}, {REPEAT as needed for additional DVs} in {population}.” Updated 9/2018

Lab 2 – Data Presentation  4

Spring 2019

Let’s look at an example: Research Question #1: Is there a difference in hematocrit between male and female college students? Results: Table 1. Average hematocrit values for male * (n=8) and female (n=12) IPHY 3435 students. Sex male female

Hematocrit (%) (Average + SD) 47.3 + 2.1 42.2 + 3.2 Figure 1. Average hematocrit values for 8 male and 12 female IPHY 3435 students. Hematocrit was measured with a rotary tube reader. An independent ttest was used to compare the means. * indicates statistical significance at p...