I just spent my entire weekend grading 37 design reports for the thermistor lab—it has not been a fun weekend. The coming week or two will be grading hell, as I have homework due for the 72-person class Monday, Wednesday, and Friday (with another lab report due next Monday), and no grader or TA.
This lab report was the first of the quarter, so there were a lot more problems with the submissions than I expect to see on future lab reports. I’ve tried to collect some of my notes on the more common writing errors for this blog post, with the intent of trying to work them into the chapter on lab reports in the textbook:
- Some students had wordy introductions. I want reports to start with a clear, concise statement of the engineering goal, not a dump of any random factoid that might be vaguely related to the report.
- Report should be standalone—not referring to homework. If something in the homework is needed, incorporate it!
- Use paragraphs with one topic each. Every paragraph should start with a topic sentence, and the rest of the paragraph (if there is any) should support and amplify that topic sentence. It is better to have one-sentence paragraphs than to ramble from topic to topic without a paragraph break.
- Fit your model to your data, not your data to a model. You should never be changing your data to make it fit your theory—you should be changing your theory to fit your data.If you say you are fitting your data to your model, you are claiming to commit scientific fraud.
- Best-fit curves are not necessarily lines—students don’t have a “line of best fit” in this lab, because the models we’re fitting are nonlinear.
- Figure captions should be paragraphs below figure, not noun phrases above figure. Any anomalies or interesting features of the figure should be pointed out in the caption. Most of the crucial content of the report should be in the figures and captions, because that is all 90% of readers ever look at in a science or engineering paper.
- Refer to figures and equations by number, rather than “schematic below” or “equation above”.
- Don’t use screenshots for schematics or gnuplot output—export graphics properly as PDF files and incorporate them into the report so that they can be printed at full resolution even when scaled.
- Many students use way too much passive voice. Using passive voice is a way to hide who did something or deny responsibility (see Nixon’s “mistakes were made”) and should not be necessary in a design report.
- Use past tense for things that have been done, not present tense. Also, “would” is not some formal version of the past tense—it is a marker for the subjunctive mood in English, which has a whole lot of different uses. In technical writing, the most common use of subjunctive is for “contrary to fact”. If you say “I would put the thermometer in the water”, I immediately want to know why you don’t—I expect to see the sentence continue with “, but I won’t because …”
- “Software” is an uncountable noun, which means that it can’t be used with the indefinite article “a”. There are a lot of uncountable nouns in English, and there isn’t much sense to which words are countable and which aren’t—even closely related languages with similar notions of countable and uncountable nouns mark different nouns as uncountable. I’ve only found one dictionary that marks countability of English nouns—the Oxford Dictionary of American English, which is available used for very little money.
- Equations are part of a sentence (as a noun phrase), not random blobs that can be sprinkled anywhere in the paper. No equation should appear without a textual explanation of its meaning, and the meaning of its variables.
- There was a lot of misuse of “directly proportional” and “inversely proportional”: A directly proportional relationship plots as a straight line through zero. The voltage output in the thermistor lab is not directly proportional to temperature—it is increasing with temperature, but the function is sigmoidal, not linear. Similarly, an inversely proportional relationship between x and y is a direct relationship between 1/x and y. It plots as a hyperbola. The resistance of a thermistor is not inversely proportional to temperature, as the resistance is proportional to
not
.
- Read the data sheet carefully! A lot of students claimed that their thermometers were good to 150°C, but the data sheet said that the thermistor they were using had a maximum temperature range of –40°C to 105°C, not 150°C.
- Students need to use the right metric prefixes. For example, “kilo” is a lower-case “k” not an upper-case “K”. This becomes even more urgent for “micro” (µ), “milli” (m), and “mega” (M). At least one report needs to be redone because the students claimed a value around 200MΩ, when they (probably) meant 200mΩ. What’s a factor of a billion between friends?
- Some students are clearly not used to using the prefixes, because I saw a lot of values around 0.0001kΩ, which should have been written 0.1Ω (or even 100mΩ). Even worse, a lot of students just wrote 0.0001, with no indication what the units were (that triggered a number of “redo” grades on the reports).
- “Lastly” is not a word—”last” is already an adverb. The same goes for “first”, “second”, and “third”. Perhaps it is easier to keep this in mind if you think of “next”, which is in the same class of words that are both adjectives and adverbs. For some reason, students never write “nextly”.
- The × symbol (\times in LaTeX) is only used for crossproduct, not for scalar multiplication (except in elementary school). The normal way to show scalar multiplication is juxtaposition of the variables, with no operator symbol.
- “Before” and “after” make no sense in the voltage divider circuit. You can sometimes use those terms in a block diagram that has a clearly directed information flow from inputs to output, but not for talking about the two legs of a voltage divider.
Gas station without pumps 