• Keyword Searches: Enter words and phrases into search engine to retrieve documents

• Topical Index Searches: Search alphabetical listing of general concepts and terms

• Table of Contents Searches: Browse table of contents and drill down to specific topics
  
  
  

Keyword Searches

Tabitha has an inoperable, degenerative disc in her lower back. To alleviate the pain, her doctor recommends that she slim down to her ideal weight and take daily regiments of nutritive supplements. She buys a year’s supply of supplements for $3,500 from a health food store. Can she obtain any tax benefit for her expenses?

Also, Alfredo uses “doctor,” while Billie apparently understands that deductibility hinges on whether a physician prescribes the vitamins.

Finally, Alfredo uses “expense” as a search term. However, such a common word likely will not narrow down the set of relevant documents very much. In contrast, Billie selects “medical” to assure that the search request retrieves only tax authority involving medical deductions.

To conduct successful keyword searches, the tax professional must think in terms of actual words appearing in the documents he or she wishes to retrieve. For instance, when searching for judicial decisions, the researcher should select words the judge or justice likely used in writing the court’s opinion. As noted in the next section, this mindset differs significantly from that required when using an index approach.

Similarly, less common words or phrases often lead to more effective and efficient keyword searches. Common terms or general concepts may appear in too many documents and, thus, provide little assistance in narrowing the search to a manageable number of relevant tax authorities. For instance, if you wish to know when the cutting of timber is treated as a sale or exchange, the words “cut” and “timber” would narrow down the set of relevant documents much better than “sale” and “exchange.” The latter terms appear too frequently in tax authorities to be very helpful in keyword searches.

Some of the best “keywords” often are not words but numbers. The most obvious example is Code sections. Tax professionals who know the subject matter of many Code sections have a decided advantage in conducting keyword searches. In the preceding example, a researcher who remembers that Code Section 213 deals with the medical expenses deduction might select “213” as a search term. (Code sections differed under the pre-1954 statute, so use section numbers cautiously when you think older cases and rulings may be relevant to your research.)

Developing a list of keywords does not assure an effective search since relevant documents may use similar but different words. The researcher must identify synonyms and similar expressions also. Omitting a synonym can result in a missed document that is pertinent to the research issue. However, consistently identifying all synonyms is difficult (except for the most erudite wordsmiths). Thankfully, most electronic legal services provide a thesaurus. Use the thesaurus to identify keywords similar to the ones you have already chosen.

Many search engines require phrases to be entered in quotation marks while others do not. Check the reference guide or help link for each service before starting your search. Failing to enclose phrases in quotation marks when required causes the research service to interpret your phrase as separate words that, depending on the search engine, may be connected by AND or OR. Inattention to this detail might result in either pulling a large number of irrelevant documents or missing some relevant documents. At the same time, don’t enclose a phrase in quotation marks unless you are fairly certain that all relevant documents will contain the exact phrase. For instance, a researcher entering “business deduction” as a mandatory search term will miss documents providing alternative phrasing such as “deductible business expense.” The table below shows which electronic services require quotation marks around phrases.

Until the 1990s, tax professionals formulated all keyword searches using Boolean logic. In simplest form, Boolean searches involve connecting keywords and phrases using AND and OR. However, natural language searching is a more recent innovation that even non-Booleans can use. It only requires researchers to enter queries in plain English. Relying on an underlying mathematical algorithm, natural language requests automatically convert a query to Boolean logic and conduct the search. Preliminary testing suggests that Boolean logic is superior for some searches while natural language searches produce better results in other cases. Separate lessons on this website provide more details about Boolean and natural language searches.

Topical Index Searches

Keyword approaches using Boolean or natural language systems are not the only ways to locate relevant tax authorities. Most electronic research services (LEXIS is an exception) also provide comprehensive indices that you can search. Like indices in the back of many textbooks, an electronic service’s index categorizes information alphabetically by topic rather than by the actual words appearing within documents. Thus, the mindset for conducting index searches differs fundamentally from keyword searches. Whereas keyword searches focus on the actual words appearing in target documents, index searches emphasize general terms and broad concepts that might lead you to relevant documents. Researchers must think like the editors constructing the index rather than the authors of sought documents.

Index entries usually do not contain an exhaustive list of words and phrases that you might use to formulate effective and efficient keyword searches. Stated differently, words and phrases that you might include in a keyword search request often do not appear in an electronic service’s index. Indices commonly include many general terms and broad concepts that researchers would not find useful in keyword searches. So, tax researchers entering indices searching for very specific words or phrases often will be disappointed with the results. The proper mindset is to begin with a general term or broad concept to locate a main heading within the index. Researchers often find more specific terms as secondary headings beneath a relevant main entry. For instance, consider the following index headings for the main entry, “employee deductions”:

Employee deductions

Education

Books and supplies

Conventions

Transportation and travel

Tuition and registration fees

Entertainment

Clubs and other facilities

Gifts

Meals

Tickets to events

Home office

Convenience of employer

Gross income limit

Meeting patients, clients, or customers

Principal place of business

Moving

Distance test

Duration test

House hunting

Storage of household effects

Temporary living

Transporting household effects

Travel between residences

Reimbursement plans

Accountable plans

Non-accountable plans

Transportation

Automobile expenses

Reporting requirements

Travel

Away from home

Business vs. pleasure

Foreign travel

Lodging

Meals

A salesperson purchasing tickets so her major customers can attend the Super Bowl wishes to know whether the ticket prices are deductible. Her tax adviser might begin an index search looking for main entries under “ticket” or “entertainment.” Depending on the index, the tax adviser may not find a main entry labeled “ticket” or even “entertainment.” If these comparatively specific terms do not provide a lead, the researcher should think more broadly. Entertainment is one type of employee deduction. So, the tax professional should begin the index search looking for “employee deduction,” which should lead to more specific topics such as “entertainment” and, finally, “tickets to events.”

Table of Contents Searches

The third approach works best for tax professionals already familiar with the area of law related to the research assignment and its major issues. Like the table of contents appearing at the front of many books, this search approach allows researchers to begin broadly and then drill deep to specific documents. The main headings may identify various databases within the electronic research service such as annotated tax services and professional treatises. Researchers choose a database and then select a chapter (or similar division) within the database. They continue in this manner through various layers within the chapter until they eventually drill down to relevant documents.

The table of contents approach fails when researchers drill holes in the wrong spots, producing “dry holes” (borrowing oil industry terminology). The remedy involves backing out of the hole and re-drilling in another location. The more familiar tax professionals are with the area of law being researched, the less likely they will drill (or re-drill) in the wrong place.

Drilling multiple “dry holes” can be frustrating. To minimize this frustration, electronic research services often allow tax professionals to begin with a table of contents approach and drill down as far as they feel comfortable. At this point, they have identified a subset of the original database (e.g., a specific chapter or section of a chapter). Then, they conduct a keyword search on this reduced database.

Multiple Approach Searches

The preferred search technique among many students seems to be keyword searching. In fact, even after they’ve seen different approaches in class, some students rely almost entirely on keyword approaches when conducting tax research. As a result, they do not locate key tax authority for some research assignments or spend too much time looking for that elusive document they know must be out there somewhere.

One simple strategy for locating all tax authority relevant to a research assignment is to use more than one approach. For instance, after conducting a keyword search, you might switch to an index or table of contents to determine whether relevant documents have been missed. Often, relevant tax authority exists that a keyword search does not retrieve. Sometimes, the failure of a keyword approach lies in the relevant authority’s unusual or unexpected way of expressing a term or phrase that was difficult for the tax researcher to anticipate. At other times, the researcher does not formulate the optimal search request. In either case, following up a keyword search using an index or table of contents approach can locate relevant tax authority that otherwise might be overlooked.

Conversely, someone relying entirely on an index or table of contents approach to tax research can miss pertinent tax authority that keyword searching would find. Publishers of electronic databases do not always classify tax authorities in the most logical portion of an index or area of a table of contents. Sometimes a key judicial decision or revenue ruling is categorized in only one area when it might be equally applicable to other areas. The researcher doing an index or table of contents search can easily overlook this authority. At other times, a researcher might select the wrong concept in an index search or take wrong turns as he or she drills down into a table of contents. A follow-up keyword search often can locate relevant authority that an index or table of contents search misses.

All electronic research services do not permit all three search approaches. The following table highlights the available options:

Selecting Databases for Keyword Searches

Rerunning the same search request over and over in different databases consumes much time, an inefficient approach to keyword searches. So, most electronic research services allow individuals to search multiple databases simultaneously. However, routinely searching all available databases can create inefficiencies too, so experienced researchers select only databases likely to retrieve useful information. Selecting the best databases for a search will improve as a tax professional’s expertise grows. Also, the best database for a given issue depends on the researcher’s expertise. Consider the following examples in which researchers of varying expertise investigate the deductibility of corporate organization expenses:

• Inexperienced tax professionals might choose to search two traditional tax services, two professional treatises about corporations, and a newsletter simultaneously. The tax services and professional treatises provide editorial commentary that may lead to primary tax authorities relevant to the issue, including statutory law. If recent changes affect the deduction of organizational expenses, the newsletter may provide helpful insights.

• Tax professionals proficient in Boolean logic and who know that the Code and regulations do not directly address the specific organizational expense issue might bypass the editorial commentary and search only databases containing cases and rulings.

• Seasoned tax professionals specializing in corporate tax law might wish to check recent journal articles and newsletters for creative insights.

Individuals using the same databases time and again often define and name the collective group for easy access in future searches. Thus, the tax professionals in the three examples above might define the aggregate databases they selected and name them “Corporate Commentary,” “Cases & Rulings,” and “Journals & Newsletters,” respectively. Whenever they run similar searches, selecting the customized databases can save time and, thus, increase efficiency.

Occasionally, individuals wish to conduct searches only within documents retrieved through a prior search (i.e., a sub-database). Some electronic research services allow these “searches within searches” or “searches within results.” For instance, LEXIS contains such a feature, which it calls FOCUS. (Similarly, Kleinrock allows you to conduct local keyword searches using “Find” within a particular document.) The two situations below show when a search within retrieved documents might save time, increasing efficiency:

• The original search request contains several general terms, and the researcher is confident it contains all relevant documents. However, the retrieved set is large, requiring considerable time to review. Searching within the search results using more specific keywords can reduce the documents requiring review to a manageable level.

• The original search request contains some keywords that are more specific (or less common) than others, at least in the retrieved set. To illustrate, assume an individual enters a search request using the terms “dividend” and “patronage.” Reviewing the retrieved set, the researcher notices that “dividend” appears frequently within most documents while “patronage” occurs infrequently. He wishes to read each document where “patronage” appears, but the electronic system takes him to each page containing either “dividend” or “patronage.” To make matters worse, some documents contain many pages, so perusing all documents takes a long time. He could search within the documents originally retrieved using only “patronage.” This search request retrieves the exact same documents but also facilitates review since the researcher can advance to the exact place in each document where “patronage” appears.