When you formulate searches with terms and connectors, Lexis and Westlaw behave similarly, but not identically. One important difference: Westlaw treats searches using only the OR or AND connectors as Natural Language searches, which means Westlaw's default relevancy search algorithm, not Advanced Search Boolean logic, generates results. Lexis uses Boolean logic to process searches with OR or AND connectors. To force Westlaw to run an Advanced Search using Boolean logic, add the adv: prefix, as here
adv: cats AND dogs
Another difference, unrelated to their search engines, is that the services will produce different numbers of results for identical searches, because the services host different materials. For example, Lexis includes some unpublished cases not in Westlaw, and vice versa.
Compare results in Westlaw and Lexis for searches in all Cases (as of June 2018):
| Search | Westlaw | Lexis |
| erisa AND fiduciary AND actuary | 457 cases | 1124 cases |
| adv: erisa AND fiduciary AND actuary | 1091 cases | n/a* |
| erisa /s fiduciary /s actuary | 119 cases | 128 cases |
| erisa /p fiduciary /p actuary | 216 cases | 245 cases |
| erisa /10 fiduciary /10 actuary | 50 cases | 60 cases |
*Remember that adv: is Westlaw's way of representing an "Advanced" search, i.e., a terms and connectors search. Therefore, it will not work in Lexis. But Lexis will perform a terms and connectors search if you toggle the menu to the right of the Lexis search field to read Terms and Connectors.
Suppose you are interested in the cases in which your terms all appear within the same paragraph, i.e., the 216 (Westlaw)/245 (Lexis) cases resulting from this search:
erisa /p fiduciary /p actuary
Specifically, you want to find cases likely to address the fiduciary duties of actuaries. Use the atleastn() operator to focus on cases that mention actuary multiple times. Further, to include references to all forms of the term actuary (actuarial, actuaries), use the exclamation point operator.
| Search | Westlaw | Lexis |
| erisa /p fiduciary /p actuary AND atleast5(actuary) | 70 cases | 74 cases |
| erisa /p fiduciary /p actuary AND atleast3(actuary) | 107 cases | 117 cases |
| erisa /p fiduciary /p actuary AND atleast5(actuar!) | 95 cases | 101 cases |
| erisa /p fiduciary /p actuary AND atleast5(actuar!) AND atleast5(fiduciary) | 82 cases | 87 cases |
Keep in mind that this comparison only addresses how using the connectors and operators changes the numbers of results. Other than the obvious fact that Lexis produces more results using these terms, we can't draw conclusions about whether or not the cases resulting from these searches are superior to either service's default Natural Language search and its attendant relevancy rankings.
Additional examples, adding the phrase connector:
| Search | Westlaw | Lexis |
| erisa /p fiduciary /p actuary AND "fiduciary duty" | 196 cases | 188 cases |
| erisa /p fiduciary /p actuary /p "fiduciary duty" | 119 cases | 94 cases |
| erisa /s fiduciary /s actuary /s "fiduciary duty" | 52 cases | 29 cases |
As above, we can't explain without further investigation why these Westlaw results now exceed those on Lexis. Still, the relative changes in numbers of results from search to search make perfect sense. We would expect there to be fewer cases in which all of our terms appear in the same sentence as compared to where our terms appear in the same paragraph.