SCIENTIFIC METHODS
OBJECTIVES: By the end of this laboratory period you should be able to:
- Define science.
- Explain the importance of variables and controls in scientific experiments.
- Define scientific theory.
- Write a basic scientific research report.
- Compute a standard deviation.
- Graph scientific data.
INTRODUCTION
The word science is derived from the Latin "to know" and Webster's Dictionary defines it as knowledge covering general truths or the operation of general laws as obtained and tested by the scientific method. Scientists investigate the natural world via inductive reasoning, that is, they attempt to discover general principles by careful investigation of specific cases. The application of these general principles to the investigation of other specific cases is called deductive reasoning. The methods of science are usually presented as a series of steps, the scientific method, which each scientist must methodically follow: recognition and formulation of a problem, collection of data through observation and experimentation, and the formulation and testing of hypotheses. Although most scientific investigations eventually include these processes, there is no single scientific method which all scientists follow. Every scientist is different, every scientific problem is unique and each problem may require a variety of different approaches. There may in fact be as many "scientific methods" as there are scientists.
Most scientific investigations involve loops through a variety of observations, literature reviews, pilot studies and ideas from colleagues which result in the formation of a working hypothesis. Once a working hypothesis is formulated it may enter another series of loops in which its validity is tested. An experiment is a test of a hypothesis. Often a variety of factors or variables may influence the outcome of an experiment. Sound scientific methodology usually requires a series of experiments in which only one variable is examined at a time. Observed differences between these experiments and a parallel control experiment reveal the influence of the variable(s).
Experiments are followed by data analyses which result in rejection, reformulation or conditional acceptance of the initial hypothesis. Hypotheses are often conditionally accepted until independently verified by others or confirmed by further investigation. It is important to note that scientists attempt to disprove hypotheses, not prove them. Support eventually comes from accumulated inability to discredit.
A scientific theory is a hypothesis supported by a great deal of evidence which stands the test of time, often tested and never rejected. Note that the general public uses the word theory to imply a lack of knowledge or a guess, just the opposite of the scientific meaning. Darwin's theory of evolution by natural selection is an example of a scientific theory. It has survived scientific scrutiny for more than 130 years and evidence to support it continues to accumulate.
SCIENTIFIC WRITING
(from Brower, Zar & Ende, 1989)
Just as there are common methods to scientific research, scientific writing has a fairly uniform style which reflects the investigative process. At a minimum, scientific reports contain at least four sections: Introduction, Material and Methods, Results and Discussion.
1. Introduction Section
In the introduction of the paper state the nature of the problem, objectives of the study and any hypotheses to be tested. Also, give a brief background for the study, which would typically include a brief review of the literature. Relate the problem and its significance to the general discipline of study. This part of the paper presents the background, justification, and relevance of your study.
2. Materials and Methods Section
Procedures in research reports are usually detailed enough for the reader to have an accurate idea of what was done in the study or to be guided to appropriate literature for this information. A good description of materials and methods used is one that would enable a reader to duplicate your investigative procedure. Keep to a minimum the details of standard and generally known procedures (such as how an item was weighed). In a field study, a general description of the study site is called for. If this description needs to be lengthy, then it may comprise a separate subsection or a new section. Materials used in the investigation are not merely listed, as in a cake recipe, they are woven together with the methods in a narrative form.
3. Results Section
This portion of a report gives the facts found, even if they are contrary to the hypothesis or expectation. Listings of raw data are rarely presented, except occasionally in a class activity or as an appendix to the report. Instead, data typically are summarized using means, frequency tables, percentages, or other descriptive statistics for presentation and analysis in some appropriate statistical manner. These data summaries may be incorporated into figures or tables if this results in additional clarity or helps illustrate a pattern or trend. In general, the number of data collected (sample size, n) should be indicated, and some measure of variability of the data should accompany statements of means. Statistics used, type of data analysis performed, and mode of presentation depend on the study and type of data collected. Statistical comparisons of different groups of data are often called for. The results section is not just a data summarization or a collection of tables and figures; it should contain an explanation and description of the data. Tell the reader exactly what you found, and what patterns, trends, or relationships were observed. For example, do not just say "The species-area curve is shown in figure 1." Tell the reader what is being presented, "Figure 1 shows that the number of species in the habitat increases and then levels off as the area of the habitat increases." Illustrations in the results section may consist of graphs, photographs, or diagrams that visually depict your results. All such illustrations are individually numbered, cited in the text and referred to as a figure. Labeling and citing tables of data in the text is done in the same manner as for graphs. If a graph will summarize the data as well or better than a table, then the graphical presentation typically is preferable. Each figure and table should contain an explanatory legend. Be sure the axes of all graphs are fully and correctly labeled with a scale marked off and the units of measurements given; units of measurement (metric) must also be given for tabular data. Avoid the tendency to cram too much information into one graph or table, thus losing readability.
4. Discussion Section
In the previous section of the paper the results are summarized and described. In this section they should be interpreted, critically evaluated, and compared to other research reports; and conclusions should then be drawn based on the study and its findings. Whereas the results section presents the "news" the discussion section contains the "editorial."
In the discussion, examine the amount and possible sources of variability in your data. Examine your results for bias and evaluate its consequences in data interpretation. Develop arguments for and against your hypotheses and interpretations. Do not make generalized statements that are not based on your data, known facts, or reason. Be sure to relate your findings to other studies and cite those studies. Draw positive conclusions from your study whenever possible.
STATISTICAL ANALYSES OF DATA
The collection, statistical analysis and interpretation of data is an essential part of
most scientific research. Rows and columns of numbers rarely provide much useful information without further
manipulations, and it is nearly impossible to compare data from several experiments by
visual examination alone. Imagine trying to interpret the results of thousands
of experiments without some means of statistical comparison. Although there are
perhaps hundreds of different methods of statistical analyses, some of which require
complex calculations and computers, a few are relatively easy to calculate and they yield
valuable information. The range of an array of data represents the spread between
the lowest and highest value. The mode is the most frequent class or value. The median
is the middle value of an array of data, therefore half the values are greater than it and
half the values are less than the it. The mean or arithmetic average (
) is a much more useful value than either the mode or median. It is the
sum (å ) of the observed values (x) divided by the number of
observations (n).
However means give no indication of variation within a sample. Two different sets of data may have identical means if all the data are clustered very closely to the mean or if the data are markedly above and below the mean. For example, students A and B may have identical mean exam scores of 75 in Biology class but one may have individual scores of 74, 75 and 76 while the other has scores of 50, 75 and 100. The mean alone does not reflect the relative consistency or variation in their exam performances throughout the term. The variance (S2) provides an estimate of the degree of dispersion or variation within a particular sample. Variance is calculated by summing the squares of the deviations of all the observations from their mean and dividing by n-1.
The standard deviation (S) is a more commonly used estimate of the degree of
variation within an array of values. It is merely the square root of the variance: S = 
S2. Standard deviations, when used in conjunction with means,
allow different sets of data to be compared in meaningful ways.
GRAPHING QUANTITATIVE DATA
Most people find it easier to comprehend the significance of data expressed visually rather than numerically. When expressing scientific data one picture is indeed often "worth a thousand words". Scientists often use graphs to visually present data. Graphs consist of perpendicular axes, with the horizontal called the x axis (abscissa) and the vertical called the y axis (ordinate). Graphs sometimes reveal relationships between two variables, for example, one variable (the dependent variable) may be affected by the other (the independent variable). The independent variable is plotted on the x axis and the dependent variable is plotted on the y axis. If two characteristics appear to be related they are said to be correlated. For example, there is a strong correlation between smoking and lung cancer. Correlations allow scientists to make predictions when only one of the variables is known. For example, the more a person smokes, the more likely they are to develop lung cancer.
Scatter diagrams are constructed by plotting individual points on x,y axes. As the name implies, line graphs are constructed by plotting individual values on x,y axes and connecting them via lines. Histograms, also called frequency diagrams, are constructed by grouping data into discreet classes, represented by blocks, and plotting them versus their frequency of occurrence on the x and y axes, respectively. For example, numerical grades are usually grouped into the classes A, B, C, D and F. Graphs often reveal that although there may be considerable variation in the data (a wide range), most of the values occur relatively close to the mean. This is known as central tendency.
Graphs of scientific data, particularly biological data, frequently yield normal (bell-shaped) curves or distributions. Such normal distributions have the following characteristics: the mean, median and mode are identical; 68% of the total variation falls within the range of plus or minus one standard deviation; and 95% of the total variation falls within the range of plus or minus two standard deviations. Recall that standard deviation is a measure of the variation or dispersion of the data about the mean. Visually represented on a line graph, data with a small standard deviation forms a sharp, narrow peak like a mountain, whereas a large standard deviation yields a graph shaped more like a broad hill or dome.
PROCEDURE
A variety of organisms from bacteria to mammals experience periods of dormancy (diapause) between periods of activity. In all cases the breaking or emergence from dormancy is regulated by certain environmental signals, e.g. temperature, humidity, light intensity and/or duration, availability of nutrients, etc. You will be provided with various diapausing aquatic "animals" enclosed within transparent, protective coatings. Your assignment is to investigate the breaking of dormancy in these "animals" and to generate and test hypotheses which account for this activity. Keep accurate, detailed and neat records of your activities. You are especially encouraged to discuss your ideas, methodology and experimental design with your instructor and fellow students. Very few scientific investigations begin totally "cold", advice from colleagues with previous experience in the field can often make significant contributions to your work, especially during the early stages.
YOUR ASSIGNMENT
DUE SEPTEMBER (20 POINTS)
In order to provide a larger sample size and a wider array of data than that obtained in class, a web based Java applet will be used to generate data. These data will be used to write a short scientific paper consisting only of the Materials and Methods and the Results sections. The Materials and Methods section will detail your in class procedures but the data from the Java applet will substitute for the data generated in class. The data generator requires a Java enabled web browser. If your home computer does not support Java then you must use a university computer.
Use the data generator to obtain a sample size of 50 and plot the data on a scatter diagram. For this exercise the graphs must be done by hand. Computer generated graphs are not acceptable. Be sure your graph has a title and the axes are appropriately labeled. Include the raw data as an appendix to your report. Also provide a table of appropriate statistics (mean, range, standard deviation) for the emergence times (do not show your calculations).
Your report must be neat and typed, and conform to the form outlined above (this includes graph paper). It is due September 20 during your regular lab period.