Once you have decided that the resources you need are scholarly articles, and you have chosen an appropriate database from the Find Articles tab, the next step is to create a search.
Most people have learned to search using Google, but this is not efficient when using scholarly databases. Those who are searching efficiently do 4 things:
This guide is going to spend some time covering how to prepare a more efficient search strategy and some of the features of databases that make them useful for conducting scholarly research.
Preparing an efficient search strategy requires you to:
A thesis statement or a research hypothesis is not equivalent to a search you would do in a database. You may have to translate a thesis statement into multiple searches in order to understand multiple aspects of an issue.
The main things to keep in mind when creating a searchable question is that it needs to be or have:
Specific enough to your needs
Accurate and non-biased
Which of the following is a searchable question that follows SAM principles?
Example 1: Does diet improve diabetes?
This question does not follow SAM and is therefore not a good searchable question. Diet is not specific - you will get thousands of results for hundreds of different diets. Improvement is not measurable, nor is it specific, you want outcomes that you can actually measure. That requires a bit of specificity in the outcomes. Diabetes is also not specific - which type are you truly interested in? Overall, this question is not accurate for what you are truly trying to find in the literature. You will get a lot of garbage results and spend a long time cherry-picking.
Example 2: Does a low-fat diet lower blood glucose in type-II diabetics?
This question seems to solve many of the issues from above, but it is still not quite a SAM question. It is much more specific, it has a measurable outcome, and the papers you find would appear to be more precise and accurate. But can you spot the problem with this question? This question is BIASED. If you were to search it the way it is written, you would ONLY find articles discussing how low-fat diets lower blood glucose levels. You wouldn't necessarily retrieve any articles arguing the opposite, because by including directional terms like "lower," "reduce," "increase," "raise," etc., you are conducting a biased search.
Example 3: Does a low-fat diet effect blood glucose in type-II diabetics?
This is an example of a SAM question. To avoid the bias from above, we used a neutral verb between the action and the outcome.
Once you've built your question, you now need to break it apart into it's most important individual topics. Each concrete idea in your question probably has multiple ways to express it - synonyms. You need to spend a bit of time brainstorming synonyms that represent the discrete main ideas of your question. Many times you may not be using exact synonyms, but a combination of words that represent the same idea. In the chart below, you can see hemoglobin A1c being used instead of direct synonyms like blood sugar because in medicine, the hbA1c test is a measure of blood glucose.
Remember: Computers are dumb, YOU are smart. Computers do only as much as they are programmed to do and cannot truly understand human language. They don't understand that words have meaning or that synonyms exist, therefore you have to provide that information to the computer.
* Abbreviations are often not recognized as standing for a full word by computer searches. In the case of carb vs. carbohydrate, since carb is just the first 4 letters of a longer word, you may pull up results talking about both, but to be sure you capture everything available you will want to use both the abbreviation and the full term. The difference between Hemoglobin A1c and HbA1c definitely needs to be explicitly given to the computer.
** Just as computers don't understand or "know," the difference between roman numerals and arabic numerals has to be taken into consideration when searching as well.
After you have brainstormed lists of keywords and understand the main topics in your question, you then need to form a search string using the symbols and operators that databases and search engines understand.
Computers understand mathematical logic, so when you set up a search string, there are three main operations you can do. These operations are represented with three short words, known as Boolean Operators. Each operator tells the computer to perform a different type of action between the terms, or sets of terms, in your search.
Looking back at our list of synonyms, utilizing our symbols and combining ideas appropriately with Boolean operators, we might end up with a final search that looks similar to this:
("low carb diet" OR "low carbohydrate diet" OR "paleo diet" OR Atkin's) AND ("blood glucose" OR "blood sugar" OR HbA1c OR glycohemoglobin) AND ("type II diabetes" OR "type 2 diabetes" OR "diabetes mellitus")
The above search can be copied and pasted into a single search bar, like those seen on the front page of most databases.