Consumer Report
Grammar Checkers
Open Forum
Low-Cost Macintosh Printers
Report
December 1990
PC Technology Assessment Program
EPA National Data Processing Division
Information Centers Branch - RIC II, MD-35
Research Triangle Park, NC 27711
Telephone: (919) 541-0568 (FTS) 629-0568
Printed on Recyced Paper
-------
PC TAP CONSUMER REPORTS
From the Editor's Desk
As we put the final touches on this PC TAP Consumer Report, we're painfully aware of the fact that you
haven't yet seen the last one, PC TAP Consumer Report #9, which features our second report on
Microsoft Windows 3.0. That publication was sent to the print shop at EPA Headquarters in Washington
during the week of October 29. We apologize for the print shop's difficulties in producing our Report #9,
and we sincerely hope they resume their usual one- to two-week turnaround service with the present
issue.
The topic of this report, software grammar checkers, is an interesting one, especially for those whose
duties require them to write materials for general distribution. Although correctness of language might not
be of great importance for informal, internal correspondence, most authors want to be confident that
material written for wider distribution or for public scrutiny follows generally-accepted rules of good usage.
White the products reviewed in this report wont make everyone's prose sound like the work of a
professional writer, they probably can help conscientious writers avoid common grammatical pitfalls. We
hope you enjoy reading the opinions of our grammar checker reviewers.
In this issue's Open Forum, beginning on page 16, Dave Levesque describes his investigation into the
area of tow-cost printers for the Macintosh. Dave, an 1C consultant at the Washington Information Center,
has lots of experience with the Mac, and his search for an under-$1,000 300-dpi printer should be
interesting for our readers who have Macintosh computers in their offices.
The PC TAP External Resource Network (ERN) is still growing. We're gratified by the response this idea
has generated, and we're pleased with the excellent cooperation from the ERN representatives in our PC
TAP evaluation studies.. Another update on the people who make up the ERN can be found in This &
Tfcafonpage 18.
As 1990 draws to a close we are reminded of the advances that have been made in desktop computing
this year. LANs are spreading like wildfire, and Windows is the byword in the industry. Within EPA, a new
PC procurement contract offering more powerful hardware is in place, and the first NDPD combined
conference was a resounding success. It has been a year of great progress in Agency computing, and
we're pleased to have been a part of it.
The target of our first major study in 1991 is dBASE IV version 1.1. We will be looking at the viability of
both the standalone and networked versions as a data base product, and at what's involved in migrating
from dBASE ill Plus to dBASE IV. We look forward to bringing you our findings in the new year.
David A. Taylor
PC TAP CoortBrmtor
-------
Grammar Checkers
Introduction
The tact that most folks dread the task of expressing themselves through the written word is probably one
of the main reasons why those of us who like to write are kept busy. Today’s word processing packages
provide easy text formatting to help novice writers prepare good-looking written communications, and they
usually include spell checkers to enable us to elirrinate embarrassing misspellings from our writings. As
a result, it’s fairly easy to produce attractively formatted prose that’s free of rrwsspelled words, but which
also may contain atrocious grammar and makes little sense to the reader.
Enter the software grammar checker! These products are often called grammar/style checkers, because
they frequently have a built-in capability to check the “style” in which a document is written, as well as its
grammar. For example, you might indicate that the writing is intended for a “technical” audience, or that
it is aimed at readers with a specific level of education. The software then performs its checks with
consideration for the characteristics and reading ability it associates with the audience you’ve specified.
It’s easy to latch onto a grammar checker with the assumption that it will solve all your writing problems,
and that it will make you a great writer; after all, that’s what some of the advertisements suggest. This is
especially true if you have to do a lot of writing and you don’t like to do it, or if you perceive that you aren’t
very good at it. A program that promises to fix all your writing problems is just what the doctor ordered!
Unfortunately—and not unexpectedly—these programs aren’t the panacea they might appear to be.
Nevertheless, they aren’t without value to the writer who’s looking for a tool to help hone his or her
wordsmithing skills.
BenefIts of Grammar/Style Checkers
Although grammar and style checkers aren’t able to analyze a document’s content, their capabilities to
perform mechanical checks exceed those of word processing spell checkers. Not only do many grammar
checkers correct spelling, but they search documents for errors in word usage, mechanics and punctuation,
sentence structure, and subject-verb agreement. Some programs generate summary statistics indicating
the grade level of the writing, word frequency counts, and the average length of sentences contained in
a document.
Many writers report that they find these programs indispensable. Often when they’re too tired and bleary-
eyed to continue proofreading their documents, they let a grammar checker program do the work. On the
other hand, some writers associate a high level of frustration with grammar checker usage. Contributing
to this point of view is the fact that some grammar and style checkers are characterized by numerous
inconsistencies. Not only do they find a disproportionate number of false problems, but, when used
indiscriminately, they can contribute to more writing problems. For example, a poor writer who’s unable
to distinguish among the program’s inaccuracies will be unable to assess correctly its recommendations.
Grammar/style checkers are adequate proof readers, but the writer must be able to interpret and draw
conclusions about the software’s grammatical advice.
2
-------
How They Work
Grammar and style checkers have advanced significantly since they first appeared at the beginning of the
1980’s. The early programs scanned documents, one letter at a time, searching for strings of text that
exactly matched words or phrases already assembled in a software “phrase dictionary.” Like programs now
on the market, they flagged passive voice, sexist language, jargon, and mechanical errors. However, the
first-generation programs had a minimal understanding of grammar and style because they could not
capture the relationship between words in a sentence.
Today’s grammar checker software products use grammar “parsers.” Remember the sentence diagrams
you constructed in grade school? Grammar and style checkers operate in much the same way; they
compare each word in a sentence to a dictionary file listing the parts of speech of common English words.
Included in the dictionary are irregular verb conjugations, and singular and plural forms of words. Once
the program identifies each word’s range of potential functions, the parser examines the order in which the
words are placed, determines the main grammatical components of a sentence, and “diagrams” it.
Each sentence is parsed into dependent and independent clauses, and subjects and predicates. Words
are identified as articles, adjectives, prepositions, adverbs, and other parts of speech. The program checks
subject verb-agreement; if the verb is plural, the software checks to see that the subject is plural too. Then
the program might compare the text with its dictionary of objectionable words and phrases and flag those
it finds for the user’s benefit. Finally, any errors that were found are matched with an entry from a table
of messages like “No verb in the sentence” or “Word is overstated or pretentious.” This advice then is
presented to the user.
ReadabilIty Index
Most grammar and style checkers generate a readability index that indicates how much difficulty the
audience may have reading a document. Most products allow the user to indicate an audience of a general
type (e.g., business, technical, academic) or grade level (8th grade, 10th grade, 12th grade, etc.). The
characteristics of the specified audience then are taken into consideration when the readability index is
calculated.
One or more of three different techniques commonly are employed by these programs to calculate the
readability index: the Flesch-Kincaid Grade Level, the Flesch Reading Ease Index, and the Gunning Fog
Index. The Flesch-Kincaid Grade Level formula yields a number corresponding to the educational grade
level required of the reader. A Flesch-Kincaid score of 10, then, equates to the 10th grade; individuals with
less than a 10th-grade reading level probably will find the text difficult to understand. Here are the steps
to calculate the Flesch-Kincaid Grade Level index:
1. Multiply the average number of words per sentence by .39.
2. Multiply the average number of syllables per word by 11.8.
3. Subtract the sum of the values from steps 1 and 2 from 15.59 to obtain the grade level.
The Gunning Fog Index also reports in terms of grade level as an indication of the difficulty of your writing.
Based on an assumption that multi-syllable words and convoluted sentences make the document more
difficult to read, the Gunning Fog Index is calculated like this:
1. Calculate the average number of words per sentence.
2. Find the number of words with 3 or more syllables.
3. Multiply the sum of the values from steps 1 and 2 by 0.4 to obtain the Fog Index.
3
-------
The Flesch Reading Ease Index is graded on a scale of 1 to 100. The higher the number, the easier the
document is to read. Writing at the “standard” Flesch level of 60 to 70 translates to about the 7th or 8th
grade level. The following table aids in the interpretation of the Flesch Reading Ease index.
90-1 00
80-90
Easy
5th Grade
70-80
Fairly Easy
6th Grade
60-70
Standard
7th-8th Grade
50-60
Fairly Difficult
Some High School
30-50
Difficult
High School-College
0-30
Very Difficult
College Level and Up
The Flesch ReadIng Ease Index
Many federal and state agencies are beginning to mandate the use of readability indexes as a means to
ensure that critical documents can be understood by most Americans, especially where legal documents
and regulations are concerned. And use of a readability index to help authors create routine documents
is just the tip of the iceberg for this type of software. The next step in the evolution of grammar/style
checkers is on the horizon: style replicators. These programs create “style models” against which other
prose can be compared. For example, a common set of writing characteristics can be defined, and
documents can be compared with that standard. In a matter of seconds, a writer can get a report from the
software about sentences within a document that aren’t in harmony with the organization’s established
standard style, along with suggestions about changes that will bring the new document more in line with
the desired norm. For an organization, there are clear advantages to this approach to writing. With
confidence that the style model will detect deviations from the desired “look and feel” of a piece of writing
and the grammar checker will take care of linguistic mechanics, managers and editors can concentrate on
the content of a document rather than spelling, grammar, and style.
- Operating Features
While all four of the products we evaluated allow some flexibility and user control over the way in which
the program operates, significant differences among them remain. Some offer the user more options than
others. Not all of them check writing style. However, most grammar/style checkers will provide most, if
not all, of these features: -
Word Processor Selection. Some programs have a comprehensive
fist of word processors they support; others support fewer specific
prOducts, along with ASCII files.
Style Checking. Several products allow the user to specify that the
writing is intended for a specific type of audience, like “general” or
“technical;” or that it’s geared for readers with a particular level of
reading ability, like “high school” or “college grad.”
Very Easy 4th Grade
4
-------
• Rule Setting. All grammar checkers operate according to certain
rules. Some products have a greater number “rule classes”
(punctuation rules, grammar rules, style rules, etc.) than others, and
the number of available rules and rule classes vanes among products.
However, in most cases the user has some control over the nature and
complexity of the checks that will be made on a document, and will be
able to “turn off” rules that perform unwanted checks. For example,
much technical report writing is done in the passive voice. Grammar
checkers invariably encourage use of the active voice, and they flag
every instance of passive voice in a document. The software should
offer the user the option of turning off the “passive voice rule.”
• Operating Mode. Some grammar checkers offer more than one
operating mode. In batch processing, for instance, the user simply
enters the name of a text file to be checked and the name of an output
file in which the “marked up” copy should be placed. The software
runs the grammar checks, and places a copy of the original text, along
with remarks and highlighted errors, in the designated output tile. The
user then must use his or her word processor to retrieve the “marked
up” copy for review. Some programs also offer an interactive mode,
in which the software goes through the document, highlighting “errors”
in much the same way that WordPerfect’s spell checker operates. The
user has the option of correcting the error on the spot interactively,
either by selecting the software’s recommendation or by perforrr ng his
or her own edits. At the end of the interactive grammar checking
session the user may elect to replace the original with the revised
version, or to save the corrected copy under a new name.
Most users seem to prefer the programs that offer the most flexibility. Some of our evaluators were very
critical of any program that didn’t offer interactive grammar checking. They said it was too much trouble
to leave the grammar checker and go to the word processor to look at the marked up copy and make
corrections. Others said the interactive check was too tedious. These individuals don’t want to see each
error as it’s flagged, they just wani an annotated copy that they can scan to select only the revisions with
which they agree. As we’ve said in other PC TAP reports, “different strokes for different folks.” In the
discussions that follow, we’ll point out the ways in which these characteristics are employed in each
package.
Software We Tested
Many of the articles we found about grammar checkers refer to the same three products: Right Writer from
RightSoft Inc.; Correct Grammar, from Ufetree Software; and Reference Software’s Grammatik. Although
these three definitely appear to be the front runners among grammar checker products, we found another,
from Lexpertise USA, called PC Proof. All four of these products are available for the Macintosh; the first
three retain their same names, with something like “for the Macintosh added on; PC Proof becomes
MacProof in the Macintosh world. Essentially, operation of all four products is functionally equivalent in
both environments; but, not unexpectedly, procedures differ somewhat between the Mac and the PC. The
following discussion is based on using Correct Grammar on an MS-DOS machine. In the following
paragraphs, we will discuss each product primarily as it’s used on a DOS machine, with pertinent
5
-------
comments from our Mac evaluators. We’ll provide a brief overview of each vendor’s product, and we’ll talk
about how they work. Then we’ll present our evaluators’ feedback about their experiences using these
programs. The products are discussed in alphabetic order by program name; no ranking or value should
be inferred by the order in which they are addressed.
Correct Grammar
We used Release 2.0 of Correct Grammar for our PC TAP study. The $99 package will run on an IBM PC,
PC-XT, PC/AT, PS/2, or 100% compatible. Approximately 460K of free RAM is required (380K if you have
a 128K EGA or VGA video card; 4001< if you use expanded memory), along with 1.2 megabytes of free disk
space. DOS release 2.0 or later is required. The program is compatible with WordPerfect, WordStar,
Microsoft Word, PC Write, and any other word processor that can create standard ASCII files.
Correct Grammar for the Macintosh requires a hard disk, on which the program’s data folder will take up
1.2 megabytes of space. The computer must have at least 1 megabyte of RAM, with at least 580K usable.
System Version 4.2 or higher is required. Documents created with these word processors can be checked:
Microsoft Word (1, 3, and 4), Microsoft Works, Microsoft Write, QuickLetter, WordPerfect, MacWrite,
MacWrite II, WriteNow, Acta & Ada Advantage, and Appleunk. Plain text (ASCII) documents also can be
processed.
In addition to providing clear instructions about using the product, the excellent documentation includes
discussion of the philosophy and mechanics employed by Correct Grammar in the grammar-checking
process. Installation was easy and straightforward on both PCs and Macs. On the PC, special provisions
are made for the word processing programs listed in the previous paragraph. For example, the word
processor’s dictionaries will be checked for unusual words and spellings. Additionally, WordPerfect users
can install Correct Grammar in such a way that you can “hot key” from within WordPerfect to check your
documents.
Although there are references in the manual to “style,” Correct Grammar does not check style in the sense
we defined it eatlier. In the case of Correct Grammar, the settings of the various rules, rather than a slant
toward a particular audience or “document personality,” are referred to as style. However, among the
“readability” rules is a grade level check that can be set by the user.
Using Correct Grammar is easy and straightforward. Upon entering the program, the user is presented with
six options:
1. Check. Allows interactive processing of a document, stopping at each
“error” the software finds.
2. Setup. To change any of the setup options (dictionary file name,
screen colors, keyboard layout, etc.).
3. Rules. Here’s where you get an opportunity to review the grammar
checking rules, and turn off any you don’t want applied to your
document.
4. WP (Word Processor). Select this option to indicate that, for this
session, a word processor other than the one specified during Correct
Grammar installation Will be used.
6
-------
5. reaDability only. When this option is selected, only the readability
rules (long paragraphs or sentences, mufti-syllable words, grade-level,
etc.) are applied. Spelling and grammar are not checked.
6. Mark up. This is the “non-interactive,” or batch check option. Correct
Grammar saves the original document with a .BAK extension, then
performs all checks, inserts comments within brackets in the text, and
saves the marked-up text under the original file name. The user can
then use his or her word processor to review the marked up copy, and
make the desired revisions.
When Correct Grammar completes its checks, a summary like the one below is presented on the user’s
computer screen. The summary also is written to the file CG.SUM.
GRAMMP.R.BAK Correct Grammar Readability Analysts
(Also written to file CG.SUM)
2 paragraphs, average 2.5 sentences each
5 sentences, average 22.6 words each
113 words, average 5.4 letters each
184 syllables per 100 words
1 passive sentences 20 % of total
0 long sentences 0 % of total
0 misspelled words 100 % correct
0 other errors corrected 100 % correct
3 sentences hard to read 40 % correct
Flesch Reading Ease score 28.2 Very difficult
Grade level required 16
tLS. adults who can understand 5 %
Flesch—Klncaid grade level 14.9
Gunning Fog Index 17.4
User Feedback
Several of our evaluators commented that Correct Grammar imposes significant limits on the user’s ability
to move back and forth in the text when performing edits during an interactive session. For example, you
can’t scroll forward or backward more than two or three paragraphs. Correct Grammar processes the
document from beginning to end, presenting small segments in a “window” to the user. Once you’ve
moved beyond the scope of a “window,” the software can’t back up to that text segment. The only way
to access it a second time is to start at the beginning and work down again.
Favorable comments from users were related to the documentation, the help that’s available during
interactive document checking, and the appropriateness of the advice offered by the software. The table
of summary statistics also got high marks. On the negative side, some people fell that execution was slow
on a 286 PC. The fact that the software can’t be used as a desk accessory displeased our Macintosh
evaluator. On a scale of 1 to 5, with 5 being the most favorable score, our evaluators gave both the
Macintosh and PC versions of Correct Grammar an overall average score of 4.
7
-------
Grammatik
Grammatik is available in several different versions: GrammatiklV(successorto Gramrnatik III ), Grammatik
Windows (for Windows 2.x or 3.0), and Grammatik Mac (for the Macintosh). There’s even a Grammatik
Government Edition that takes government writing style into account. Our evaluators used Grammatik IV
and Grammatik Mac we at PC TAP explored the product with a copy of Grammatik Windows. The DOS
version is for IBM PCs and PS!2s and compatibles with at least 512K of memory and a hard disk. The Mac
version also needs at least 512K. Both retail for $99. According to the manual, Grammatik Windows
needs 450K of memory, but with Windows 3.0’s memory management capabilities, that’s a moot point.
On our hard disk, the Grammatik subdirectory occupies just over a megabyte.
The list of word processors supported by Grammatik is most impressive: we counted 35 specific packages,
plus ASCiI text on the word processor selection list. Grammatik also is compatible with the Ventura
Publisher desktop publishing product. No significant installation problems were reported with any of the
Grammatik versions, although one evaluator had trouble setting up the option to hot-key to Grammatik from
WordPerfect. You can even have it installed and running without consulting the documentation. An
excellent tutorial called “tour” is worth your time; not only does it illustrate how Grammatik proofs a
document, but it covers all the rule classes as well.
When you use Grammatik, its very important that you select a style from among the six categories of
writing under the “Preferences” menu: general, business, technical, fiction, informal, or custom. The
category you select determines which of the rule classes Grammatik will use to analyze your document.
The “custom” option allows you can create your own personalized style, which Grammatik will use when
checking your writing. The government edition offers a seventh style, “proposal writing,” which recognizes
a long list of acronyms common to government writing.
After the checking of a document is completed, Grammatik produces a statistical analysis of the document.
The program graphs your writing’s readability index, which you then can compare with the Gettysburg
Address, a Hemingway novel, or an insurance policy. If you have defined a personalized style based on
a particular document that you specified, the current writing also can be compared against that “standard.”
For example, if you’re a technical writer and you have written what you consider to be an excellent
technical document, you might want to define that document as your standard. The figure below shows
the readability comparison chart for our test file “GRAMMAR2” as displayed in Grammatik Windows.
FLESCH READIHO EASE SCORE
gra r2
Cettysburq Address i f 92
He ing ay Short Stor9 92
Life Insurance Policy
FLESCU—KINCAID GRADE LEUFI
gra .ar2 3
Gettysburg AddresS 3
Heutinqway Short Story 3
Life InSurance Policy 13
Readability ConfiariSOn Chart
8
-------
User Feedback
Grammatik IV prompted varied responses from our evaluators. The documentation and interactive help
were favorably rated. The hotkey option in WordPerfect 5.1 was seen an excellent option. One person
thought batch processing was preferable to interactive because it’s faster and you aren’t tied to your
computer while Grammatik does its thing. Others were impatient with responding to every “problem” the
program encountered. Some prefer interactive mode because they don’t want to have to exit to their word
processor to see the marked-up copy and make the corrections. We think the length of the document has
a lot to do with which mode is best. We found it tedious plodding through long documents in interactive
mode, but for short items, like a memo for example, interactive is fine.
On a scale of 1 to 5, with 5 being the most favorable score, Grammatik got about a 3.5 for
comprehensiveness and accuracy of grammar checking. Several people reported frustration over the
treatment of “who/whom” and “they’re/their/there.” The software allegedly recommends that the user
consider changing to whichever of these alternatives is not in use. For example, one person said it flagged
several iterations of “whom” in a document, and recommended changing them to “who.” The
recommended change was made, and Grammatik was asked to look at the document again. This time it
flagged “who,” recommending a change to “whom.”
We were unable to replicate these problems in Grammatik Windows. Although we weren’t able to find any
references in documentation suggesting that the Windows version was an upgrade, we called Reference
Software to investigate that question further. They said the Windows version was not considered an
upgrade to Grammatik IV, but that some additional rules and other improvements were included in the
Windows release. A new release of Grammatik!Vis planned for the first quarter of calendar 1991. If it’s
as good as our Windows version, it’s sure to meet with excellent reviews. As was the case with Correct
Grammar, the only serious criticism of the product on the Macintosh was the fact that it isn’t available as
a desk accessory.
PC Proof/MacProof
At $159 for the PC and $195 for the Mac, these products from Lexpertise Linguistic Software are the most
expensive grammar checkers we looked at in this study. The PC version runs on 100% IBM compatible
DOS machines of at least XT class, with 640K of RAM and a hard disk. MS DOS or PC DOS version 2.0
or later is required. In addition to ASCII text files, PC Proof is compatible with Microsoft Word 3.0, 4.0, and
5.0, and with WordPerfect 4.2 and 5.0 (although we had no problems running with WordPerfect 5.1).
On the Macintosh, MacProof is a desk accessory! This was quickly pointed Out with much enthusiasm by
our Mac users. Hardware requirements include a MacSE, MacIl, CX or Cl, and 1 megabyte of RAM.
MacProof is compatible with MacWrite 4.6 and 5.0, MacWrite II, Microsoft Word 4.0, MindWnte 1.0 and 2.0,
and PageMaker 3.0, 3.01, and 3.02 (2MB of memory required for PageMaker only). ASCII files also are
supported. Unlike the other grammar checkers we evaluated, this family of products was developed first
for the Mac, and later for the PC.
This program isn’t as powerful in some ways as the others we’ve discussed. For example, it doesn’t
support any style checking. Neither does it produce a statistical summary, although there’s an “analyze
structure” function that shows certain statistics upon request. The total number of words in a paragraph
or in the document can be displayed, as can the number of sentences or paragraphs in the document.
Averages are available for words/sentence, words/paragraph, and sentences/paragraph.
9
-------
An excellent user guide is provided, but we found PC Proof easy and intuitive to use with no help at all.
You simply start it up and follow the on-screen directions. All that’s necessary is to name the file you want
to examine, provide the name you want the corrected file to be saved under, and select the “run checks”
option from the main menu.
PC Proof checks your document a paragraph at a time. It highlights all the problems in the first paragraph,
then waits for you to decide what to do. If you want to know that’s wrong, move the cursor to a highlighted
word and press F2 (highlight); the problem will be defined in the “problem” window, and an explanation also
is presented. The figure below shows an exanqle of PC Proof in action.
PC Proof Checks. -
— Problem — Capitalization
Capitalizatxonj Any sentence which does not begin with a capital
Often Confused letter is flagged as a capitalization error.
Individual words inside a sentence are flagged as
capitalization errors if the foxm of the word in the
text does not correspond exactly to the dictionary
form.
_______________ Unfotmatted input text
I will be out of town for several days at the Cci ex show. I am flying
back Tuesday night. Could we meet either sometime Wednesday or the
first thingmursdaymorning? I spoke with Mr. Zirkle about your
suqqestion for capitalizing the software development, and he loved it.
hopefully the auditors will buy it. I ‘in anxious to get the business plan
finalized so I can get back to the Western Savings proposal.
Last but not least, have you set a wedding date yet’ Marlene and I are
very excited about your engagement to Sue. Did you know that Marlene is
best friends with Sue’s older sister Shauna’ Be sure and let us know when
32 insert =
Fl Help P3 Previous PS Next F l Ignore P9 Reprocess Tab Window
P2 High Light P4 Prey/Spot PG Next/Spot PS Oict PlO Go to Esc Main Menu
As you move through the document interactivety, you can edit the text as suggested by the program or
ignore its advice and move on. Any changes you elect to make are reflected in a revised version of the
original that PC Proof saves under the output file name you specified when you first entered the program.
It’s as simple as that.
We ked PC Proof and MacProof, and so did our evaluators. The program is well organized, and it has
attractive, easy-to-follow screens. Moving through a document interactively doesn’t seem as laborious as
with some of the other products either. One person pointed out the ability to focus on only one sentence
out of an entire document as a great advantage. These programs don’t do as many things as their
competitors, but they are very good at the functions they do perform.
10
-------
Right Writer
At $95 list, RightWnter is the least expensive (by a $4 margin) of the products covered’in this report. The
PC version will run on any PC, XT, AT, or PS/2 with at least 384K RAM and DOS 2.0 or later. The
program supports most popular word processing programs, including any word processor that will produce
an ASCII file. Word processor selection is performed easily from a list presented to the user by the
software.
RightWriter for the Macintosh will run on any Mac with 512K of memory available, operating system 3.2
or later, and Finder 5.3 or later. Supported word processors include MacWnte 2.2-5.0; MacWilte Il 1.0 and
1.1; Microsoft Word 3.0 and 4.0; WordPerfect 1.0.0 and 1.0.4; and WnteNow 1.0, 2.0, and 2.2. ASCII files
from any word processor also can be processed.
RightWriter does not have an interactive processing mode. The program is very easy to use; all major
functions are clearly listed on the main menu, which is reproduced below.
RightWriter CR) Version 3.1
Copyright 1989 by RightSoft, Inc. Licensed to: pc tap
MAIN MENU
ANALYZE DOCUMENT
REMOVE COMMENTS
CHANGE SETTINGS
MODIFY DICTIONARY
EXIT TO DOS
Check a document for proper gralrunar, writing style, word usage, and
punctuation. Get comments, overall ratings, and recommendations.
Fl for Help U to Select Option — to Execute ESC to Cancel
Selecting ANALYZE DOCUMENV yields another screen upon which the user is prompted for the names
of the input (where the document to be analyzed is) and output (where the revised text is saved) files.
After entenng the file names, simply press Fl 0 to turn control over to RightWnter; the user then is returned
to the main menu. At this point, the process can be repeated for another document, or the user can exit
RightWnter and examine the marked-up file with the word processor.
When RightWriter finishes the analysis of a document, the user is presented with an abbreviated analysis,
as shown at the top of the next page. A more detailed analysis is appended to the marked-up document
that RightWriter places in the output file specified by the user. The summary created during the analysis
of a brief test document is reproduced on the bottom half of page 12.
11
-------
RightWrfter (A) Version 3.1
Copyright 1989 by RightSoft, tnc Licensed to: pc tap
RIGHTWRITER P NALYSI S
Analyzing Document: d: \wp5l\dave\grammar2.
Marked—Up Copy: d: \wp5l \dave\grammar2 . OUT
Word Processor: WordPerfect
Number of Words in Document: 90
Number of Unique Words: 59
Number of Words Analyzed: 90
Readability Index: 3.00
Strength Index: 0.69
Descriptive Index: 0.18
Jargon Index: 0.00
You may now use your word processor to view, edit
or print your document.
Press Any Key to Continue
— — *5> >
Overall critique for: d:\wpSl\granunar2.
Output document name: d:\wp5l\grarrinar2.OUI’
IIA I3XLXTT : 3.00
4th 6th 0th 10th 12th 14th
1* I I I I I I I I
S LZ I —
P..ad.r. n..d a 3rd grad. level of .duc*tice.
PT H : 0.69
0.0 0.5 1.0
I****I****I * ***I*** 5I****I****I****I I
Tb. strength of delivery 1. good, but can b. i ,zov.d.
DEB IPT1W I : 0.10
0.1 0.5 0.9 1.1
i* I t t I I I I I I
TWZ I I WOODY
Tb. use of adj.utiv.. and adverb, i. no al.
- ‘ UD : 0.00
5seTa a . r i Z C D XOOS:
7. Mont s.ntenc.s begin with pronomes.
Try using other .enteao. start conditions.
13. Yew prepositional pbr s.. are ua.d.
<>
Review this list for negativ. word. (0), argcu (J),
colloquial words (C), niau.ed word. (0), ai.sp.lling. (7),
or words which your reader ny not undaratsnd (7).
nobody (0) 1 _____ query (7) 1
<< OO WOODS TO Pmvi.w LIST>>
RightWriter Document Summary
12
-------
User Feedback
We received mixed reviews on RightWriter. Not unexpectedly, those who prefer batch processing like
RightWriter, while users who like interactive grammar checking are less enthusiastic about this product.
Everyone agreed, however, that RightWriter’s method of always saving the marked-up copy with a different
extension could likely cause disk clutter. To preclude accumulating an endless stream of files with an .OUT
extension, the RightWriter user must be very conscientious about cleaning off such files after their
usefulness has ended. This problem isn’t unique to RightWriter, however. Grammatik and PC Proof create
a backup copy of files they process, and Correct Grammar offers the user the opportunity to do so before
it saves the revised text over the original contents of a file. Summary files containing statistics and other
details also require housekeeping to keep them from eating up disk space.
All our evaluators agreed that RightWriter’s documentation was excellent. The product’s user friendliness
seems to be its strong suit. The screens are well laid out and easy to read, and the options are clearly
presented and intuitive. One evaluator commented on the usefulness of the report of summary statistics.
There was some disagreement about the quality of the grammar checking done by RightWriter. One
person reported that it missed a number of “obvious errors,” while another said this product caught obvious
grammatical problems that another grammar checker missed. One evaluator commented that the marked-
up document created by RightWriter was hard to read. Another liked the marked-up copy, and pointed out
that it was far easier to read the annotated hard copy than to “grind through” an entire document
interactively.
As with the other products included in this report, preference for RightWriter seems to be a personal choice.
Overall, averaging feedback from users of both the Mac and PC versions, RightWriter gets a grade of 4.5
on the 5-point evaluation scale.
Summary
In our introduction to grammar checkers on page 2, we noted that these products aren’t cure-aIls for
untrained or inexperienced writers. The question of whether or not you should invest a hundred dollars
or so of your software budget in a grammar checker is an issue everyone must consider individually. The
amount of writing one does certainly is a major consideration. If you write only a couple of memos a month
for distribution to people within your office, then a grammar checker probably isn’t worth the investment in
terms of the money to buy one, or in time to run your memos through the program. On the other hand,
if you do write extensively and produce documents for external distribution, you rr ght find it worthwhile
having a grammar checker in your software repertoire. As one of our reviewers commented, “It’s certainly
easier as a writer editor to use the grammar checker to confirm your suspicions about someone else’s
writing! But when they argue with you, it’s another way to enhance your credibility!”
When one scrutinizes the evaluation feedback we received on the grammar checkers that were the subject
of this report, no clear “winner” emerges. All four products have strong points, while none does a flawless
job of what they’re designed to do: checking that the grammar in written material conforms to generally
accepted rules of English usage.
On the other hand, all have the potential to be really useful writing aids for the individual who already has
an understanding of basic grammar. You may say “Oh, great. . . if I already know basic grammar, why
do I need a grammar checker?” The answer is, we need them for the same reason we need spell
checkers: to make sure we don’t carelessly violate any of the grammatical rules with which we’re familiar.
These packages are also excellent at helping a writer stay within the parameters of a particular writing
style, or to avoid vocabulary or construction that may exceed the reading capabilities of the intended
13
-------
audience. So the greatest benefit of this type of software may be in its ability to help one avoid common
pitfalls and write for best effectiveness, rather than as a grammar tutor.
It’s important to note that grammar and style checkers are still an evolving software category. We
mentioned on page 9 that Reference Software is preparing to release a Grammatik update in the Spring,
and a new release of Correct Grammarwas announced within the past couple of weeks. According to the
announcement from Ufetree Software, Correct Grammar 3.0 will have improved error checking, a style
guide that includes eight built-in style categories, an expanded grammar rule set, and a 20% speed
increase with a decreased RAM requirement. We’ve ordered the upgrade, and will include a brief review
of it in a future PC TAP Consumer Repoit.
We have collected reviews of grammar checkers from a number of trade publications dunng 1990. The
list, in alphabetic order, includes Bay Area Computer Currents, Conputer Reseller News, InfoWorld, PC
Hands On, PC Magazine, and MIS Week. In addition, PC Week published a grammar checker feature in
the September 1989 issue. Most of the artides we’ve seen report on three packages: Correct Grammar,
Grammatik, and RightWriter. Usually the publications did not rank the products, but several did. PC Week
tested the “big three” and scored them like this on a 10-point scale: Grammatik III (8.8 points), Correct
Grammar 1.0 (8.3 points), and RightWriter 3.1 (6.6 points). (Note that these were not the versions of
Grammatik and Correct Grammar that we used in our study.) MIS Week evaluated Grammatik (3½ stars
out of a possible 5) and RightWtiter (3 stars). On the 5-point scale used by PC Hands On, the overall
ratings were Grammatik, 3; both Correct Grammar and RightWriter, 2.
Feedback from participants in our PC TAP study indicates that each of the four products reviewed in this
report has its strong points, and each has characteristics or lirntations that some users probably will feel
they can’t live with. Fortunately, in today’s marketplace one usually can find a source that will allow you
to try a product before you buy, or that will sell software on a 30-day money-back satisfaction guarantee.
Grammar and style checkers can be a great help to the serious student of writing. We recommend you
try one—or all of them—and select the one that best suits your personal needs and preferences.
14
-------
Study Contributors
PC TAP appreciates the efforts of the following people in testing software and contributing to the content
of this report.
Vicki Dougherty
National Enforcement Investigations Center
Denver Federal Center
Denver, CO
Susan Frank
Public Affairs Office
EPA Region I
Boston, MA
Sophia Jeffnes
Information Consultant, RIC I
NDPD Information Centers Branch
Research Triangle Park, NC
Barry Parks
ADP Coordinator, LVF
Office of Radiation Programs
Las Vegas, NV
Dr. Nancy B. Pate
Pollutant Assessment Branch
Office of Air Quality Planning and Standards
Research Triangle Park, NC
Glenn Piper
Technical Support Specialist
American Coastal Industries
Research Triangle Park, NC
Pam Worley
Technical Writer/Editor
NDPD Publications Section
Research Triangle Park, NC
Dr. Robert Zucker
NSI Technology Service Corporation
EPA Environmental Research Center
Research Triangle Park, NC
15
-------
Open Forum
Open Forum provides an opportunity for users to share with others their own
innovations, or the results of their own technology assessments. The PC Technology
Assessment Program neither verifies nor endorses the contents of Open Forum items,
but we are pleased to offer them as a service to users.
Low-Cost Printer Options
for the Macintosh
This prmter study was submitted by Dave Levesque, of the Washington
Infottnaflon Centei staff at EPA Headquarte . Dave s an IC consultant who
specializes in Macintosh support, and in helping users with graphics
app cabons. He may be contacted at FTS 475-7418. PC TAP appreciates
Dave’s coneibution.
The objective of our study was to find a printer for the Macintosh capable of 300 dots per inch (dpi) output
for under $1000. Three Hewlett Packard printers—the DeskJet Plus, the LaserJet lIP, and the
DeskWriter—are priced under the $1,000 limit. The DeskJet Plus and the DeskWnter employ inkjet
technology, while the LaserJet is a four-page-per-minute laser printer. We decided to evaluate these three
devices for use with the Macintosh.
The DeskWriter has an AppleTalk interface that works with any Macintosh computer. Both the DeskJet
plus and the LaserJet have a parallel interface that’s intended for use with an IBM or compatible computer.
To connect the parallel printers to the Macintosh, we used a device called the Grappler. The Grappler is
available for about $60 from Orange Micro Systems. There are two models, the LX for parallel printers and
LS for serial printers.
WorkIng with the Grappler
The Grappler is a cable with a standard Macintosh 8-pin DIN connector on one end, and a small box and
36-pin centronics connector on the other end. The Mac 8-pin connector can be plugged into either the
modem or printer port on the back of the computer. The small box on the centronics side of the cable has
four dip switches that must be set to match the printer in use. A label on the box shows switch settings
for most of the popular printers.
The Grappler comes with software, including an installer, print drivers for a variety of PC printers, the Apple
imageWriter LQ printer driver, and three character fonts. Virus detection software should be turned off
before installing the Grappler software. The installer prompts the user for the type of PC printer in use, and
automatically installs all drivers and fonts. Depending on the computer system configuration, two
installation options are available. “Complete” installation is recommended for systems with a hard drive
and sufficient RAM memory (1 Meg of RAM may not be enough if there are many fonts and/or inits running
on the system). “Minimum” installation minimizes the number of Grappler software drivers and fonts
installed; we found it unsatisfactory in operation, and we used the “complete” installation for all our printer
tests. After the installation is complete the lmageWriter LQ driver must be selected in the Chooser, then
the computer must be re-booted. During normal operations, the Grappler can be managed through the
Control Panel. It can be turned on and off or configured, and options can be selected for print spooling,
font check, PC printer selection, precision font scaling and keming, and printer RAM.
16
-------
When a print command is issued with the Grappler active, a dialog box pops up offering three print quality
choices: Draft, Faster, and Best. “Draft” mode will not iandle graphics, and it offers no advantages for
printing from the Macintosh. With all our test applications, the bit-mapped images produced by “faster”
mode for both text and graphics were noticeably ragged. However, “faster” mode usually produced a page
of output in two minutes or less (about half the time required for “best” print mode). The “best” mode
caused the most problems with the Macintosh. The print quality was poor when fonts other than the three
supplied from Orange Micro were selected, and in some applications the file would refuse to print or the
Mac system would lock up. Print quality was best when one of the three Grappler fonts was used.
Generally the Grappler worked as advertised, and it appears to be an excellent low-cost solution for printing
from a Mac to most PC parallel printers. Several shortcomings are worth mentioning, however. We had
a problem placing the computer and printer on a desktop within the constraints of the Grappler’s four-foot
cable. Additionally, PC Printers connected to a Macintosh via the Grappler can’t be networked; we had
to disable TOPS (Mac network software) to prevent system conflicts. It is also important to note that the
Grappler can not print PostScript files directly. However, two images we sent from postscript programs
were converted and printed on the PC printers. The manual for the Grappler was brief and easy to follow,
but incomplete in several areas. We found it necessary to call Orange Micro in California to get the
DeskJet up and running.
Printer Test Results
The HP LaserJet lIP personal laser printer, priced at just under $1,000, provided the poorest results of the
printers tested with the Macintosh. The amount of desk space required for the LaserJet lIP also created
some problems, especially with the short Grappler cable. Additionally, paper handling was poor and
clumsy. In its defense, the lIP was easy to set up, and it can be upgraded with more memory, a PostScript
font cartridge, and an AppleTalk interface to make it a true Macintosh PostScript-compatible laser printer.
As tested, the lIP had only 512K of RAM and required the Grappler interlace cable and software.
Insufficient printer RAM caused printer errors, and sometimes the Mac system bombed when printing large
files, especially in “best’ print mode. Printing to the lIP was often frustrating when after five or ten minutes
the printer displayed an error message while the computer continued to send information. Attempting to
cancel the print operation by pressing “Command+Period” on the Mac often caused the system to lock up.
The HP DeskJet Plus, the least expensive printer in our test group at around $635, also required the
Grappler interface cable. Setup was easy, but a phone call to Orange Micro was required to obtain the
proper dip switch settings for the printer. The DeskJet has two sets of dip switches located below the
paper tray. Normally they are all set in the down position, but with the Grappler switch #5 in both banks
“A” and “B” must be in the up position. The DeskJet produced the poorest print quality, and it was the
slowest of the printers tested. Although most of the test pages printed in under four minutes in “best”
mode, complex graphics took as long as ten minutes. Unless one of the three Grappler fonts was used,
text was ragged even in “best” mode. Graphics with large dark areas came out of the printer wet, and they
smeared if touched.
With its AppleTalk interface, HP’s DeskWriter, which was designed for the Macintosh, turned in the best
overall results. At just under $800, the DeskWriter’s price includes four printer fonts. Total setup time,
including software and font installation, was less than 15 minutes. The DeskWriter can be used with a
single Mac or networked and shared by several computers. Print times even for complicated graphics was
faster than any of the other printers tested. Prints were clear, and text looked good for all applications
tested in “best” mode. “Faster” print mode produced nice quality prints in most graphic programs, but fonts
appeared a little ragged. Cabling and placement of the DeskWriter is easy due to the AppleTalk interface.
Without a doubt, this printer is the stand-out winner among those tested.
17
-------
This & ihat
In This :That we.hnn you the att st d & opments in areas we hdve
r ported on b ki é, t U you about anriounc er of int rest ng new
products or p c nt other tim I i’ ms of intorost
PC Technology Assessment Program
External Resource Network Update
The PC TAP External Resource Network (ERN) is made up of people in the Regions and Labs who
are interested in technology assessment, and who are willing to serve as local contacts for PC TAP
at their respective sites. Although we included some changes in ERN personnel in our last Consumer
Report, we want to provide a more recent update. Vicki Booth, our former Region X representative,
has moved on. Also, three more organizations have asked to join the ERN: the Environmental Criteria
and Assessment Office (ECAO) in Research Triangle Park, and the laboratories in Gulf Breeze,
Florida, and Grosse lie, Michigan. Here are the new representatives at these four locations:
Site Representative
Region X Kent Hargrave
EPA Region X
1200 Sixth Avenue
Seattle, WA 98101
Grosse lie Robert Buckley
USEPA—Emergency Response
9311 Groh Road
Grosse lie, Ml 48138-1697
Gulf Breeze Russ Ryder
EPA Environmental Research Laboratory
Sabine Island
Gulf Breeze, FL 32561
RTP Becky Mangum
Environmental Criteria and Assessment Office
EPA Environmental Research Center, MD-52
Research Triangle Park, NC 27711
-------
How to Submit Items for Open Forum
In keeping with the PC Technology Assessment Program’s objective to have the user community actively
involved in TAP projects, users are encouraged to submit items for inclusion in future PC TAP Consumer
Reports. If you have independently investigated the capabilities of a software product or a hardware
component, we would like to hear from you. We’d also like you to share with others your solutions to any
problems you may have encountered with a particular application or device, and about tricks, shortcuts,
or unique applications you have devised. Although we can’t promise to publish every contribution, we will
evaluate them all in terms of their potential interest to our readers and their conformance to the spirit and
intent of PC TAP.
There are no additional rules for Open Forum contributions, but here are some guidelines:
1. Contributions must be typed. Our first preference is that they be
submitted on a floppy disk in WordPerfect format. If that isn’t
possible, the next best method is to EMAIL the text to PCTAP,
EPA30647. The least preferable method, but still acceptable, is to
mail a typewritten article to TAP at the address on the cover of this
publication.
2. The length of your contribution will be determined somewhat by its
complexity. However, keep in mind that we’re primarily interested in
the purpose of your study project and how pleased you were with the
results, not in the nitty-gritty details of how you did it. We will publish
your name, address, and phone number for those who want more
details. Two to three pages is probably a reasonable maximum
length. On the other hand, a paragraph containing a nugget that may
be useful to others would be equally welcome.
3. All material submitted by users is subject to our editing, and you will
not be given an opportunity to review the final manuscript before
publication. Sorry, you’ll just have to trust us. If we have questions
or don’t understand any part of your text, we’ll contact you for
clarification.
We hope you enjoy PC TAP Consumer Reports, and we look forward to hearing from individuals who
have insights or discoveries to share with others. Thanks for your interest and your participation in the
PC Technology Assessment Program.
19
------- |