iFlipr is one of the leading general purpose flashcard applications for iPhone/iPod which offers interval study and the recommended graded slideshow approach. I see great potential for this application. The clean and powerful web counterpart, in particular, is impressive, and the web centered approach may indicated a general direction for applications in the future.
My review below points out many strengths of this application but also points out some issues with the flashcard interface which will frustrate high-volume students, as well as some of the limits of interval study which will concern long-term students.
Note: This review is from the perspective of language learners, and especially those who will be engaged in high-volume and long-term study of vocabulary. See the Terms page for an explanation of the technical terms used in these reviews. See the Basics page for a list of basic features found in flashcard applications useful to language learners.
The iFlipr iPhone/iPod application is complemented with an online flashcard database at iFlipr.com where, with a free account, users can download shared sets (much like StudyStack, iFlash Deck Library, Flashcard Exchange, and other similar services), create and share their own sets, and freely export sets in the CSV format. However, it might be more accurate to say that the website is complemented with an iPhone/iPod application…
A Flashcard Library Browser?
iFlipr is best described as a “client” of the web page, focused on providing an environment in which a user can download cards and practice them on their iPhone/iPod. This approach has some strengths but also may produce frustrations for and ultimately be too limited for the high-volume long-term student of language. With one exception (ironically the “Downloads” pane) all the panes within the application involve interaction with the website and require an internet connection. That means the very UI of the application is deigned with the online interactivity chosen as the primary focus of the user experience within the application. Instead, I believe it would serve the users better, and serve iFlipr’s chances in future competition with other similar applications, if it focused on the study experience itself, which does not require online connectivity.
The “Featured” pane shows you a selection of high quality sets found on the iFlipr online service. The “Recents” pane shows you recently shared sets, the majority of which I found to be of very mixed quality. The lack of a rating system, at least for the time being means that one faces something of a jungle in wading through the online offerings. The third “Search” pane allows you to search for sets to download, while the “More” pane uses a built in browser to load the iFlipr website and provide access to most of its features. These include the ability to create decks, using a somewhat more basic (compared to the excellent desktop browser experience provided by the website), but still solid input interface.
Only the “Downloads” tab is an offline component of the application, listing downloaded sets and allowing you to move into study mode.
The advantage to this approach is easily apparent for any new user to the application. They can easily find and download sets in their field of interest through the excellent interface. They can also, if they have an internet connection, quickly create new sets, or download sets they have created online via the desktop.
However, once you have your flashcard files in the application, however, all of these features become superfluous. Once you have purchased your furniture, you don’t need a whole furniture store camped out in one’s living room, as it were.
I would recommend a complete redesign of the user interface to focus on the study experience. Combine all the online interactivity into one pane of the application. Let users enter that “Online” pane and access all the above mentioned areas such as featured, recent, etc. sets and the ability to interface with the website through the mobile client. That will free up all of these other buttons for greater offline functionality to address some of the problems and missing features mentioned below.
In my interaction with the developer, we seem to have similar views on these issues, and the origins of design of the application are clearer to me after learning that this dates back to the application’s pre-SDK development as a web-based application for the iPhone. I understand that the application is still in a transitional stage and I look forward to seeing how it will change.
Creating and Editing Sets
While it can be done either on the desktop or through the built in browser on the mobile iFlipr client, sets can only be created with an internet connection and, it should be remembered, a functioning iFlipr server. On at least one past occasion, the iFlipr has had connectivity issues, which reminds us that when we become dependent on web services, our data is in the hands of a service provider. In this case it is not a major corporation with dedicated server monitors but a free and well-designed service in the hands of a, for the time being at least, an engaged developer.
One major problem for students of Asian languages or who wish to keep verb conjugation information, etc. in a separate field is that, like Mental Case, iFlipr only supports two fields. Those students will probably want to consider other options like the upcoming iFlash Touch, iAnki, iCards, or other offerings that support three fields or more.
For those who can do with only two fields, however, the browser based set creation is, however, beautifully done. A full WYSIWYG editor is provided in the web based editor which allows you to do a great deal of customization of colors, fonts, sizes, and other formatting.
These translate beautifully once downloaded to iFlipr, putting most competitors to complete shame in this area. You may also add sound and images to your cards. What’s more, through the “Settings” tab, one can customize the font and size displayed on the iPhone and the font sizes are done relative to the size indicated through the set creation editor online.
The built in browser in iFlipr also allows you to create sets, but without the rich editor, which is a wise decision. In both cases, however, it is extremely easy to move to the next card. Like the desktop iFlash application, you can simply tab between fields and it will automatically create new cards when necessary. This is in contrast with the awkward Command-N or Command-Return shortcuts required in Mental Case or Anki desktop clients. Creating, editing, and deleting cards is simply a delight in iFlipr. Its dependence on being online for the creation process, however, will be a problem for students who, and I speak from long years of field experience here, find themselves in a grimy dormitory room or hole-in-the-wall cafe in some foreign country without a (or with a very slow) internet connection and want to type up their vocabulary.
There is no way to organize or move cards between or into multiple sets (decks) though the developer has indicated on the Facebook group for the application that groups and folders are in development. There isn’t even, in fact, any easy way to get an overview of what cards are in a set unless one is editing the set through the web page interface. There should, at least, be a way to get a quick list overview of what cards are inside.
The heart of any flashcard application is its flashcard study. For the high-volume long-term student even the smallest issues here can be enough to give an elaborately designed application the toss.
There are some aspects about flashcard study that I hope the developer will give serious consideration too. Flashcard study should be fast, clean, and go easy on the hands. Many of us will be studying a hundred or more cards a day so every little moment and movement counts.
The first problem is the positioning of the buttons. Like many developers to the iPhone/iPod who treat the environment much as it was a desktop environment, they forget that thumb location is key to placement of UI objects. In this case, the very bottom of the screen is the worst location. One must stretch one’s thumb when holding the device with one hand and this repeated motion (over a hundred times, for example) will eventually lead to serious strain. The thumb can reach the upper half of the screen with much greater ease. As I have said in other reviews, the Lima Sky Kanji application approach is good: make the whole card touchable, immediately flipping upon a single tap almost anywhere on the screen. Also, I recommend making a second tap mark the card correct (instead of having to aim one’s thumb at a small check mark) while keeping the “wrong” button as a separate button. As long-term students know, a higher and higher percentage of one’s cards will be marked correct if daily study continues so it should be the default (or easier) choice.
Fortunately, iFlipr does not make the mistake Mental Case does: when marking a card correct or incorrect, it immediately moves to the next card. Well, not actually immediately. There is a slow visual flip animation which ensues. Like I have said in reviewing other applications with this problem, “It was cute the first few times…” but after a few hundred flips it becomes akin to having your computer issue a, once hilarious, barf noise when ejecting a floppy disk (if anyone remembers that craze). This same problem is found in iCards, Mental Case, and some other iPhone/iPod based programs. This seriously slows down flashcard study for high-volume students. At least offer the option of turning these transitions off.
It is also not immediately obvious (until the content is read) what side of a card one is looking at. Mental Case offers a very nice feature of showing a slightly different background shade for each different side. iFlash Touch (at least the current beta version) shows the field name in barely visible text in one corner. Either of these methods, but preferably the first, gives the user immediate visual feedback on what side they are on. This may not seem like a big deal, since one should be able to tell from the text quite quickly what side one is looking at but it can definitely make the study experience more pleasant.
A number of settings are available during flashcard study. Here one can “Reset Mastery Levels,” choose between the graded slideshow and multiple choice approaches to flashcards, of which the former is much better for serious students. They can also choose the order for displaying fields, and whether to use the interval study (Leitner) method or a simple first to last order. It would be best to offer a setting here, perhaps on this very same option, to show the cards simply shuffled. The font and font size percentage (relative to the size set when the card was created) can also be set here in the settings.
This settings window, including the “Keep Practicing” and “Done Practicing” buttons has a very web pagey feel to it, instead of that of an iPhone application but regular buttons. The UI might be improved in future by using more native iPhone controls and following the consistent look of other iPhone applications here.
Also, these settings, might better be placed in the settings for a given set, rather than all crammed in settings available only during actual study. These and many other features, such as more statistics, set management, and various other features that might be added later are good candidates for replacing the currently heavy web-interactivity nature of the panes in the home screen of the application.
Cycle elimination exists only to the extent it is part of the interval study process. Like Kanji Flip, the interval study approach is continuous. One never “completes a round.” In order to give users a degree of in-progress feedback on newly studied material, generally I feel it is best to provide a “debriefing” screen after all newly introduced cards have been introduced showing the user, at the completion of a round, how well they performed in the round, before continuing to remove correct cards from the stack. Anki provides this when scheduled cards are completed. There are ways of doing this right without a traditional cycle elimination, but it depends on the effectiveness of the interval study features. Let us take a look at this in our consideration of interval study in the application.
iFlipr immediately puts itself in the running with the more powerful applications when it chose to incorporate interval study, the availability of which is apparent when one opens the “Settings” and sees the “Leitner” method of flashing there by default.
You can see that interval statistics of a sort are being compiled through the “Card Mastery” number. This increases each time you get a word correct and resets to zero when you get it incorrect. I think reseting the word mastery is a suitable brutal punishment for the forgetful and can probably serve well in most circumstances. However, my personal experience over the years of my own interval study suggests that words that have reached any level higher than 4 or 5 that one gets wrong after many weeks or months without being prompted to review it, recover very quickly and once they have recovered don’t need to be reviewed as frequently as a fresh card at a mastery of 0-3. I thus think an optimal system is one which drops the mastery level by some number (or provides such an option) perhaps arbitrarily set by the user according to their own needs (with a default of a drop of 2 or 1 point of mastery).
The “Deck Mastery” number, which shows the lowest mastery of any card among those in a set is not a useful number and I would recommend replacing it with a more useful statistic (See the stats section of my Basics page for some possible replacements). Words are not all created equal. Not only do we find some more difficult to memorize than others, we also unconsciously prioritize all words we learn according to how useful we deem them to be. The only exception to this is in preparation for a vocabulary examination in which the probability of a given word being tested is equal to that of all other words on a vocab list but the many language programs I have gone through suggests that students quickly learn that this is usually not the case. If my set contains 100 Chinese legal terms, for example, and I have real difficulty in remembering 2 very obscure contract terms that I’ll probably never have any use for, Knowing how well I know (on average, for example) the other 98 of them is far more useful than how much I suck on two words I’ll never use but am too lazy to delete from the list.
The actual algorithm being used is not transparent to the user or even posted on the web page. Interval Study in software applications has been around since Piotr Wozniak designed SuperMemo for DOS in 1987 and I designed Flashcard Wizard for Mac OS 9 in 1999. The maturity of this approach, which all the most powerful applications now include, is such that developers could benefit from exposing their algorithms (at least in general terms) for advanced users of flashcard study to help guide them in their purchasing decisions.
The iFlipr developer Joseph Kumph is, however, open about it and was kind enough to explain to me how it is done in the case of iFlipr:
The algorithm is fairly simple: cards in the deck are broken up into different groups based on their card mastery level, and one of these groups is choosen based on an exponential decay function coupled with a random number generator. Specifically, the cards with the lowest card mastery are the most likely to be choosen. There is also a “ghost” pile, where a card goes immediately after it is marked incorrect, and kept there for about 30 seconds, to make sure the cards are not shown again too quickly.
There are a number of good elements at work here. The use of a “ghost” pile is an excellent advanced feature that will be familiar to Anki and Mnemosyne users. The iFlipr approach also guarantees that for large sets of cards, a user will be more likely to be prompted with unfamiliar words before more well-known ones. However, iFlipr has a very serious case of the Insatibility Flaw since it continually drags out words which are not on the verge of forgetting. I hope that the developer will revisit his interval study approach and develop an algorithm which focuses as much as possible on only prompting those words in need of review (with an optional study on demand feature for crammers).
It doesn’t seem that, as in the case with synching iAnki, Mental Case, or some other powerful iPhone applications, that interval study data is synched with the server. That means if you practice your flashcards directly through the web interface, this study has no connection to whatever progress has been made on the mobile client. Given the tight integration between the mobile iFlipr client and the web page, I find this somewhat unusual.
There are a few UI issues with iFlipr which may be easy to fix. Much of the application feels more like a web page than a native iPhone/iPod application. Double clicking (by mistake) ends up slightly zooming the “page” which shouldn’t happen. The panes should be solid and stick in place, but instead one can often accidently “drag” the page out of position or find it out of position when a new dialog set of options appear:
These are very minor UI problems I think can be addressed without too much difficulty.
It is fantastic that data is completely portable in iFlipr, through its online interface. There is excellent import for CSV and tab-delimited data and it will also export CSV data (without formatting data, except for for carriage returns).
The support page for the application is essentially through Facebook. Here the Wall shows that the application has some very enthusiastic fans. However, though facebook is popular, contrary to common perception, not everyone on the world is on the site yet. I would recommend providing some support interface or forum outside the environment of Facebook.
Fool’s Final Words
iFlipr is a clean and strong contender in the iPhone/iPod flashcard application market. The web interface is clean, simple, but powerful, and provides an easy way to download one’s data. However, there is too much dependence on the web interface and the client ought best gather the web interactivity options in one place to allow for gradual expansion of other useful features, especially in the realm of set management, study statistics, and so on.
As explained above, biggest issue for long-term students of the application is the Insatiability Flaw in its interval study approach. The biggest problem for high-volume students is the UI of the flashcard study itself. I thus suggest further refinements in the algorithm to eradicate this problem, and possibly provide filters for truant users to easily return to study while prioritizing words with certain tags or at certain interval stages.
The slow visual flip transition is cute the first few times but slows high-volume study significantly. I have also not tested performance of the application with the high-volume sets (3000-8000 words) that students in intensive language programs will want to subject their software to. I was not comforted by suggestions on the support page to keep it to 100 words per set. The lack of full touch-ability for the card and the hard to reach buttons in the bottom of the screen should be given a second luck in future releases.
Finally, support for only two fields will mean that students of Chinese, Japanese, etc. will want to look elsewhere for a more flexibility three+ field solution.
There are some real gems in this application, however. With some improvements here and there to address the issues above, I feel like this application could easily catapult itself to the top of the pack.
Import: Tab-delimited and CSV.
Non-Roman Scripts: No problem
Modes of Study: Graded Slideshow
Media and Frills: Images and Sound.
Entry Creation: Excellent if two fields are enough.
Entry Editing: Excellent
Set Organization: None.
Flashcard Study: Poor UI, slow
Interval Study: Fair, Insatiability Flaw
Design and Feel: Fair, could use some improvements
Golden Coxcombs: 6/10 but with great potential…