 Writing reports Writing



Writing reports

The ability to write effectively is a vital part of communicating the results of laboratory experiments. After all, if no one knows what you did, then it is not of much use.


Reports for this course are simplified “formal” reports. They will consist of the parts typically found in chemistry journals: title, abstract, procedure, data, results, discussion. Each of these parts serves a distinct purpose in presenting the results of an experiment. The title of a report should be a concise description, in the most general terms, of what was done. For example, one of the experiments in this manual involves the qualitative analysis of a set of metal ions. The title “Qualitative Analysis” is not descriptive enough to convey the scope of the experiment. Where as, “Qualitative Analysis of Common Metal Cations” indicates both what was done and what type of sample was used. Lastly, the title “Fun with Ions” offers almost no information. An abstract is a brief statement about the experiment, which summarizes the major methods and findings. The length of the abstract is only a few sentences; about 3 sentences should be sufficient for all of the experiments in this manual. Typically, the abstract is written after everything else, so that the important conclusions can be included. Below is an example of a well written abstract. 1

Transient ionic and atomic emission-absorption measurements on a train of positionally-stable copper spark discharges are described. The measurements are made as a function of discharge current duration at constant amplitude and rise time. Spatial observation zones are isolated in the spark gap with a set of masks and a relay lens to estimate the movement of electrode vapor. Absorption measurements required custom circuitry for a pulsed hollow-cathode backlight and synchronously gated dual-channel boxcar integrator. Results indicated that electrode vapor moves primarily along the interelectrode axis in response to current duration, with substantial ionization remaining in the post-discharge period. Neutral atoms remain in the gap for long times after current cessation.

Although longer than the abstracts that you will write for this class, the example contains the parts important to any abstract. The phenomenon being studied is described in the first sentence. The type of measurement is explained with attention to any unusual methods. In the last two sentences, the general results of the experiment are summarized. Notice that there is very little specific detail. 1 Because the procedures given in this manual will generally be followed as written, you are normally not required to write a procedure section in the reports. If you followed the procedure in this manual exactly, reference it in a single sentence (e.g. "The procedure described on page__ of the T. Araki and J. P. Walters. Spectrochimica Acta B. 1979, 34B, 371-383.

laboratory manual was followed."). If you deviated from the procedure, describe what you did. If you had to design a procedure, describe it in sufficient detail so that it could be repeated from your report. The data section of a report is an organized record of what happened during the experiment, including numerical data and qualitative observations. Pre-formatted data pages are provided at the end of each experiment to make this summary easier. These pages should be completed after the experiment (not during), as a way to summarize the information recorded in your notebook. Also included on the pages in this manual is a section to summarize the results of the experiment. As the name suggests, this section summarizes the calculated results of the experiment and presents any graphical data such as calibration curves. The final section in a report is the discussion or analysis. In this part, conclusions about the meaning of the data are made and the significance of the experiment is analyzed. Conclusions should be supported by direct reference to results. If your results are inconclusive and conclusions cannot be drawn, state this. If possible, compare experimental to literature values where appropriate. The discussion is also the section in which experimental errors should be explained. In most of the experiments in this manual, there are questions listed in the discussion section. These are intended to get you to think more deeply about the experiment. Simply answering these questions is not sufficient for the discussion; you must also summarize your results.


The results and discussion sections of a report are used to summarize and explain a set of data. When it is numerical data, there are some conventions for how to work with it. Generally, replicate data should be taken through the calculation process independently, so that averaging the replicates only occurs at the last step. The following example demonstrates this for two samples made for a chloride analysis. Sample 1 mass = 0.5052 g chloride concentration = 0.01389 M Sample 2 mass = 0.5122 g chloride concentration = 0.01403 M ×volume ×atomic weight of Cl = ÷ 0.5052 g Percent chloride by weight = 24.37 % ×volume ×atomic weight of Cl = 0.1244 g ÷ 0.5122 g Percent chloride by weight = 24.28 % To make an average value () for the amount of chloride in the samples, the average of the percent chloride by weight would be used. This is because only the percent weight value is a relative measurement; it is independent of the size of the sample. All of the other data is dependent on the sample size (e.g., the measured concentration is larger for the heavier sample #2). Had the individual chloride concentrations been averaged, the data would be meaningless because of the dependence on sample size.

In the preceding example of percent chloride, all of the measured values (mass and concentration) have 4 significant digits. This number of digits was carried through to the final result. To determine if that makes sense, calculate the absolute error (E) between the two samples.


 24 .

37 %  24 .

28 %  0 .

09 % The error is all in the last digit; therefore, it is correct to report the final results and the average to 4 significant digits. Notice that the rules for significant digits agree with this more detailed analysis.


Explaining the errors in an experiment is often one of the most frustrating steps in the writing process. The section on error in chemical measurements (see page Ошибка! Закладка не определена.) gives the necessary background for quantifying error and for trying to identify the sources of error. However, the only way to clearly identify sources of error is by taking thorough notes during the experiment. Records of mistakes made during the experiment are invaluable when trying to write the discussion. Whenever possible, the best way to discuss the error in an experiment is by using statistics. Presenting the relative error and standard deviation of an analysis makes the data more meaningful. Even when it is not possible the make statistical analyses, the rules of significant digits allow some judgments about error to be made. In the example above, calculating the absolute error allowed us to verify that using 4 significant digits was justified and provides a crude estimate of the error associated with the analysis. But where does that error come from? Given the types of error (indeterminate, determinate, gross) some judgments can be made based on one’s data. For example, if all of the data are too high, there would be a determinate (systematic) error in the analysis. This might be the result of miscalibration of an instrument or a consistent mistake by the experimenter. If the source of error was indeterminate, one would expect the error in the data to produce both high and low results. Being able to recognize the type of error helps to eliminate some of the possibilities, but the specific source must still be identified. Particularly in the case of gross errors (e.g., spilling some of a sample), the error should be identified, but the anticipated results must also be indicated. In an acid-base titration, for example, one might titrate past the end point. This error would be recorded with the titration data so that it can be explained in the discussion section. The error should not only be indicated, but the effect on the calculated result also be noted. In over-titrating, too much titrant is added which leads to an erroneously small calculated concentration. Hopefully, the calculated result matches that prediction. Thus, the relative direction (up or down, high or low) of an error can be predicted by looking at the calculations. When writing about such an error, determine if the predicted direction matches the calculated result. In the case of the over-titration, a low calculated concentration matches with the predicted direction of error, thus, the result is explained. If the calculated concentration were too high, then over-titration can not be used to explain the result because the direction of the error is opposite. When deciding the sources of error in an experiment, one should not speculate. Sources of error are not things that could have happened. Rather, they are events that did occur. This is one more reason why a well maintained laboratory notebook is important.


Writing for the sciences is very much like writing for any other discipline. Good writing is characterized by clarity, grammar, and logic. Clarity is always important, but especially so when writing about detailed, complicated experimental data. Be sure to say what you mean. The use of proper grammar is as important in the sciences as it is in literature. Improper use of language makes a report difficult to read and understand. The conclusions and claims made in scientific writing must be supported by experimental evidence. Statements and conclusions are supported by data, not feelings or opinions. If your data are inconclusive, it is better to state this than to force the data to fit some hypothesis. Other aspects of scientific writing:  Know your audience. For this course assume that the reader is a college chemistry student such as you.  Clarity, grammar, spelling and brevity are important. Write in the active voice when possible. But note that the topic of the writing is the data, not you.  Do not use the first person pronouns (I or we).  Use abbreviations and specialized terms when necessary, but define them the first time that you use them.  If you must start a sentence with a number, spell it (as in three or fourteen).  When writing decimal numbers (e.g., 0.1), always include a leading zero – the zero to the left of the decimal point. Numerous medical associations require this practice since failure to do so has lead to accidental deaths. 2  Learn how to use your word processing software to make your documents look professional. Most programs will allow you to construct tables without having to use the tab key, and will allow you to insert symbols such as ° or ±, and to subscript and superscript without using ^. The exception to this is calculations. Due to their complexity, it is easier, faster, and much more effective to write sample calculations by hand.  Use the words “precise” and “accurate” correctly. Accuracy refers to how close your value is to the standard or known value. Precision refers to how close together your results are; data cannot be precise unless you have done more than one trial.  Use the words “clear” and “colorless” correctly. Water is clear and colorless. Sunglasses are clear and green. Milk is opaque and colorless (white). Muddy water is opaque and brown.  Do not include statements of opinion. For example, stating that the experiment was difficult or tedious is your opinion and does not belong in a reporting of the results of your study. 2 “Hospital Is Fined $11,000 in Death of Boy.” New York Times 7 October 1998, late ed.: B6. and “Simple Error Killed Baby.” The Herald (Glasgow) 9 March 1999, p. 4.

 The literature value of a physical property is just that; it is not the "literary" value.  Regardless of what your spell-checker says, “absorbency” is not a word unless we are analyzing diapers or paper towels. Always use “absorbance” to describe measurements of light absorption.  Give properly formatted references when using other people’s results whether it be a physical constant, a value from the scientific literature, or another student’s result. (e.g., as is seen in this manual: J. Koryta, Ion Selective Electrodes, Cambridge University Press, London, 1975.)


A well-made plot is very effective in presenting data that can not be easily explained in the written form in the discussion section. However, a poorly made plot can make a report confusing and look bad. Often, the most concise way to present data gathered in an experiment is to plot it. The first challenge in doing so is deciding which axis will contain which type of information. The x-axis is always assigned to the independent variable. The y-axis is always assigned to the dependent variable. That is, the y-axis variable changes as a result of a change in the x-axis. For example, if the concentration of a species was measured over a period of time, the concentration will be y-axis values and the time will be x-axis values. This is because the concentration changed as a result of time passing. The convention of naming would describe this example as a ‘plot of concentration versus time’ (implicitly, x-axis as time and y-axis as concentration). The axes of the plot should be labeled with the parameter being plotted followed by the units in parentheses. The number of sub-divisions (tic marks) on the axes should be sufficient to make it easy for the reader to identify the value and should be scaled in easy-to-read increments. The minimum size for a plot is about half of a standard sheet of paper. The individual data points should be clearly plotted. Data from similar experiments may be plotted on the same axes if the data points are represented with different shapes and a legend is used. If only one set of data is plotted then no legend should be used. Often a line is fit to a set of points. However, this must be done with some forethought, rather than simply connecting the plotted points. A line that best fits the data should be used. The shape of this line will require some knowledge of the relationship between the x- and y-axis values. Thus, if a linear relationship is predicted from theory, then a linear fit to the data should be used. they are. Lastly, the figure, graph or plot should have an appropriate title. Restating the axis labels is not an appropriate title. A useful title indicates the overall purpose of plotting the data, not just what type of data

Figure 1: A typical plot of absorbance versus concentration. The equation for the best-fit line (linear regression) is included on the plot area.