In my review of RootsMagic 4 Consistency Checks, I mentioned that I encountered some problems during the Preview and Beta.
By the time it was released, RootsMagic 4 reported exactly the same problems as RootsMagic 3, and that still includes some false positives. During testing, RootsMagic 4 reported many more errors because the underlying date arithmetic was not working right.
Throughout my email exchange with Bruce about these issues there never was any hint that anyone else had reported this issue. Perhaps that was just professionalism on his part, but it gave me the impression I was the only one reporting this issue.
It made me wonder: were many using it and not reporting the defects, or was almost no one using it? I got the impression that Bruce was dependent on just my defect reports and that worried me. Were all those other Preview testers not taking advantage of consistency checks?
I asked Bruce and he kindly provided a few numbers. I was the only one to report the logic errors in the consistency checks (and these will be fixed in some future version), but he received multiple reports of the broken date arithmetic. Not many though, no more than 4 or 5 reports per thousand users.
The Preview and Beta ran for several months, and many trying out the RootsMagic 4 Preview were experienced RootsMagic 3 users already, yet less than half a percent of them reported that the consistency checks were unusable. That sure is not many.
Similar observation about users not using the feature have
certainly been made before, but we have some actual numbers now.
Alas, the percentage of users reporting does
not directly translate to a percentage of users using the feature. Not everyone reports defects, and
the feature was so obviously broken that many who did try it out may have assumed
that others already reported the issue.
If we assume that one of every twenty to twenty five users who noticed it went on to report it, we still have to conclude that only one in ten users bothered to take advantage of consistency checks during the Preview period. You would need to do a large test with a so-called instrumented version of the program to be sure, but Bruce Buzbee shares my impression that few users take advantage of this feature.
Consistency checks are an important feature of genealogy software. You can write a genealogy in a word processor, but one of the basic advantages of genealogy software is that it understands genealogy and can take advantage of the computer’s ability to perform many checks quickly.
That capability protects you against publishing obvious nonsense, such as people born after their burial - but only if you bother to use it. The existence of many online genealogies containing various impossibilities make it clear that many genealogists are not using it.
One reason that people are not performing consistency check is that their genealogy software does not offer these checks.
FamilySearch PAF, Legacy Family Tree and RootsMagic offer consistency checks, but GenoPro 2007, Ancestry.com Family Tree Maker 2009 and MyHeritage Family Tree Builder 3 do not. Family Tree Maker 16 had a Data Errors Report, but this feature was discontinued in its official successor, Family Tree Maker 2008 and Family Tree Maker 2009 did not bring it back.
Another reason is that people do not know about it. It may be mentioned in the manual, but how many people still read manuals? If it isn’t in the Quick Start Guide, many users will never try the feature.
Consistency checks are not a nice-to-have advanced feature, they are a must-have fundamental feature that separates tools from toys.
Vendors need to treat consistency checks as the fundamental feature it is.
Consistency checks are not a nice-to-have advanced feature, they are a must-have fundamental feature that separates tools from toys.
Vendors that offer the consistency checks should clearly highlight this in every feature comparison matrix. Vendors whose product include this feature hardly seem to promote it.
Consider Legacy Family Tree. Legacy Family Tree has the most extensive consistency checks
I have seen yet, yet their feature
comparison matrix only mentions that Potential problems are individually
selectable
. Consistency checks are too important a feature to get such a minor
mention.
Millennia highlights its Research Guidance and related features such as the research log, but that Research Guidance is going to make some pretty weird confused suggestions if the genealogy is not consistent to start with. Research Guidance is a nice feature, but consistency checks are fundamental.
RootsMagic has a feature list that mentions the Interactive problem
list near the
button of the Tools
section, which is the one-but last section on that
page. Could they downplay it even more? That it provides consistency checks should be under General Features
, near the
very top of the page. Placing it near the top helps to convey that this is a
fundamental must-have feature.
That vendors fail to highlight basic features, even though they stand to gain by it, as some of their major competition lacks those features, seems a bit weird, but perhaps not so surprising after all.
It probably reflects that their users have simply not been clamouring for consistency checks, but have been begging for customisable fonts, box styles, colours and gradients for their ancestral charts.
That users ask for more charting features does not mean they do not care about consistency checks. Au contraire, they are likely to be quite upset when they notice that the wonderful looking poster-sized chart they just received from a printing service - for more money than their genealogy software cost them - contains obvious nonsense that they had not noticed before…
The same users who did not ask for the feature are likely to consider the
software to be defective for not warning them.
That users do not ask for a
feature does not mean they do not want it, nor that they do not expect it. It
does not even mean that they will not blame the vendor for allowing them to
enter inconsistent data.
Users should not need to state basic requirements, vendors must make sure their product just works right.
After a few decades of personal computer software, many users simply assume
that modern software gets the basics right, and software vendors cannot simply
dismiss the user’s great expectation as unreasonable.
Almost everything you buy nowadays meets all kinds of basic criteria you do not
know about, and do not want to think about, yet you would upset if it did
not entirely work as expected. As users, we all have come to trust
specialist vendors to provide what we really need without us needing to spell
that out; That’s their expertise anyway, not ours. We rely on their pride in
product, market forces and perhaps some oversight to ensure a good product, even
when we have no idea what makes a good product.
The challenge to vendors is to meet all user expectations, even if those expectations are unstated. Users should not need to state basic requirements, vendors must make sure their product just works right.
Education takes many forms. Books, magazines, speakers, blog post, reviews, vendor demos, even press releases, they all inform and educate us about genealogy and genealogy software.
But how many introductory books remark upon the importance of consistency checks? How many vendor demos highlight this feature? How many reviews remark upon the absence of consistency checks when a product lacks this important feature? How many press releases highlight the speed, quality and ease of use of the new or improved consistency checks?
It is tempting to fall into the trap of thinking a feature isn’t important if almost no one is talking about it. It isn’t a sexy feature, it isn’t as immediately attractive as a large technicolour wall chart - but that wall chart may not go over well if it contains obvious nonsense.
It isn’t very hard to improve the current situation. Vendors can show consistency checks as a fundamental feature in their feature lists, highlight it in their feature comparison matrix and mention in their Quick Start Guide.
Reviewers can mention the absence of the feature in their reviews, editors can demand their reviewers pay attention to such basics. Authors can mention the benefits in their books and blog posts. Editors could go as far as disqualifying genealogy software that lacks consistency checks from their comparative reviews. Textbook that suggest software to beginners can make sure to only suggest software that does include this fundamental feature. Everyone who thinks it is important can do something to make sure this feature gets the attention it deserves.
Many of us do perhaps pick up more good and bad genealogy habits from the software we use then from the books we read or training we attend. Vendors that publish software without consistency checks do the genealogical community a serious disservice by thus allowing and encouraging a bad habit.
I sometimes get the impression that vendors whose product does include consistency checks are disappointed that so few users take advantage of it, and that these vendors feel discouraged by that.
The issue they have encountered is that just including the feature isn’t good enough. Using that feature has to become a user habit, part of the regular use of the program, and vendors can make it so by making the feature part of the regular genealogical process. There are many ways to weave consistency checks into regular use.
It is fairly straightforward to remind users to do a consistency check.
Programs that offer tips can include it as regular tip, perhaps even as one that
appears more often if the user has not performed a check in a while.
It is also possible to prompt users to perform consistency checks in the same
way they are prompted to make backup and check for application updates.
It is perhaps best to try to not just detect inconsistencies, but prevent them from occurring in the first place.
Several applications perform consistency checks when a users import data. Most genealogy applications should do that as a matter of course.
Another way to prevent inconsistencies is to perform checks as the user is entering data. Many applications already do so, but it may be worthwhile to have another look at this code, make sure it checks as much as it can, and work on keeping it user-friendly.
Most importantly, genealogy software should to prevent users from
embarrassing themselves by publishing an unchecked database, prevent users from
passing errors on to others.
Protect users by
warning them if they are about to export, print, or otherwise publish any part
of database that did not pass a consistency check since it was last updated.
Copyright © Tamura Jones. All Rights reserved.