Element data window

The Element data window can most conveniently be viewed by pressing F5 (Window : Layout : Warrior layout). This arranges the statistics and element parameters windows in the classic way first introduced in SuperMemo 3.0 in 1988. When you double-click the element data window, you can view the element's repetition history.

This is a typical element data window after a repetition:

The caption of the element data window displays the element number and the element title. In the presented example, the element is listed at position #48421 in the collection and its title is "genet: How can Tay-Sachs disease be detected?". Element titles are automatically generated from the first question text.

Element data (left panel)

Element data is displayed in the left panel of the element data window. It includes the following fields:

  • Repetitions - number of repetitions of the displayed element. If the element had been forgotten, the number of memory lapses is displayed in parentheses. Once the element is forgotten, the count of repetitions begins from scratch (i.e. Repetitions equals 1 again). In the example presented in the figure, the item has been repeated 8 times and has not yet been forgotten (no lapses shown in the Repetitions field). As the picture was taken immediately after the 9th repetition, the field will be updated as soon as the repetition data panel on the right is cleared
  • Interval - current interval of the item, i.e. the number of days between Last repetition and Next repetition and the previous interval in days. Here the item have last been repeated in 1997 (8th repetition) and the interval was 2673 days (over 7 years). The interval between the 7th and 8th repetition was 1709 days (almost 5 years). Important! If you advance a repetition, its interval and U-Factor will both be corrected for spacing effect. Consequently, the previous interval field here will point to an earlier date than the actual date of the repetition. This earlier date will equal the date on which the repetition would have to have been made in the optimum schedule to produce the same combination of memory stability and retrievability as the actually executed advanced repetition. This virtual repetition date will come earlier than the date of the advanced repetition, but later than the date of the preceding repetition. For more see: Algorithm SM-11
  • Last repetition - date of the last repetition of the item. This field will also inform you how many days have passed since the last repetition ("d.a." stands for "days ago"). Here the 8th repetition took place on March 28, 1997 and the current repetition which has just taken place is recorded on July 22, 2004. The 8th repetition took place 2673 days ago, i.e. exactly as indicated by Interval (usually these two numbers are the same only on the day of repetition)
  • Next repetition - date on which the next repetition of the item should take place, and the number of elements scheduled for repetition on that particular day. Next repetition of this item should take place on July 22, 2004 and it has actually just been made. 10th repetition has been scheduled on December 20, 2013 as indicated in the same row in the repetition data panel on the right. The presented item is only one of 4210 elements scheduled for review on July 22, 2004. This high number indicates either a longer break in learning or a heavily overloaded learning process, which may be a norm in incremental reading
  • A-Factor - A-Factor associated with the currently displayed element, and the number of times the element has been postponed. A-Factor is a rough measure of item difficulty and an accurate measure of the speed with which inter-repetition intervals will increase. The higher the A-Factor, the faster the increase in intervals. For items, the most difficult items have A-Factor equal to 1.2. For tasks and topics, A-Factors equal the increase in interval in a single review and may often be much less in value than it is the case with items. Note that Difficulty (below) is much more an accurate measure of item difficulty (as perceived by the user). A-Factor of 4.672 in the example above indicates that the items is relatively easy to remember, which can also be concluded from the fact that there has been no lapses in 9 repetitions. The item has been postpone once with Postpone
  • U-Factor - U-Factor and the current estimation of element's retrievability. U-Factor is the quotient of the previous interval and the next interval (in items that have been repeated only once, U-Factor equals the first interval). U-Factors make up an important element of the SuperMemo Algorithm. If you do not know the algorithm, U-Factors do not have much meaning to you. Retrievability corresponds with the probability of correct recall of the item at a given point in time. Theoretically, retrievability should decline exponentially from 100% on the day of the last repetition, to 100% minus the forgetting index on the day when the repetition should take place. Here U-Factor is 2673/1709=1.564, and the retrievability is 89.7%. Had the item not been postponed once (see A-Factor), the retrievability should stand at 90% on the date of repetition as the requested forgetting index had been set at 10%
  • Forgetting index - planned probability of forgetting the item at repetition (in percent). Forgetting index can be changed to a desired value (e.g. in Element Parameters). For example, if the forgetting index is 10%, you stand a 90% chance that you will remember the item during the optimally scheduled repetition. Here the forgetting index has been set at 10%. See also: Forgetting index
  • Future repetitions - estimated number of repetitions of the item in the next thirty years, and the time needed for executing repetitions in that period. This value is easily derived from A-Factor, Repetitions, Forgetting index, and the matrix of optimal factors (see: SuperMemo Algorithm). You can click on the Forgetting index field to change the forgetting index and see how that changes the estimation of future repetitions. SuperMemo roughly predicts that there will still be two repetitions of the presented item in the next 30 years. As the 9th repetition has just taken place, the most likely number of repetitions before 2034 is two, of which one should take place in 2013. Avg Time in Statistics makes it possible to estimate that the cost of retaining the presented item in memory until 2034 is 16 seconds on the assumption the item will not be forgotten in the meantime. Still, the probability of a memory lapse before 2034, assuming no delays, is 19% (two repetitions with the forgetting index of 10% result in 0.9*0.9=0.81 probability of recall) 
  • Ordinal - ordinal number associated with the element. Ordinals can be used to sort items in the pending queue, final drill queue, etc. The lower the ordinal, the higher the priority of the item. The presented item shows the ordinal 2197 (as set by the user). You cannot say if this number is high or low. It all depends on the ordinals of the remaining items in the collection
  • Difficulty - difficulty of the displayed element estimated on the basis of the following parameters: Interval, Lapses, Repetitions, A-Factor, and First grade. The theoretical minimum for the difficulty is 0% and the theoretical maximum is 100%. This number decreases gradually with successful repetitions or increases with memory lapses. In a typical collection, the difficulty of items usually ranges from 16% to 64%. If the difficulty reaches beyond 65% you should have a closer look at the formulation of your items (e.g. memory interference, etc.). The presented item is estimated to be at 21% difficulty which indicates it is relatively easy to remember
  • Delay - repetition delay as compared with the optimum date. Due to a single postpone, the presented item has been delayed by 62 days resulting in optimum-to-used-interval quotient of 1.04
  • Type - type of the element: item, topic or task (see also: Topics vs. items) and its current status: dismissed, pending or memorized. The presented element is an example of a memorized item

Repetition data (right panel)

To understand repetition parameters displayed on the right of the element data window you may need some rudimentary knowledge of the SuperMemo Algorithm. Here are the fields of the repetition data in element data window: 

  • Repetitions - number of repetitions of the displayed item (including the just-made repetition). If the item had been forgotten, the number of memory lapses is displayed after the colon. The number in the parentheses indicates the number of repetitions that the item would need to reach its current interval assuming the current value of the matrix of optimal factors and no memory lapses on the way (the so called repetition category). This hypothetical value is used to index the matrix of optimal factors and the matrix of retention factors in computing the new values of individual entries at repetitions. The exemplary element above have just been repeated for the ninth time and has never been forgotten. Due to the relatively long interval, the repetition category is quite high: 11.4
  • Optimum interval - optimum interval the item should use to ensure the forgetting probability determined by Forgetting index. The optimum interval before the next repetition is 3391 days (or over 9 years)
  • New interval - new interval before the next repetition. New interval might optimally be equal to Optimal interval; however, two factors may make these two values differ: (1) minor interval dispersion is needed to avoid scheduling a large number of repetitions on the same day (interval dispersion also speeds up the convergence of the optimization algorithm), and (2) some constraints imposed on the new interval may make it impossible for it to equal Optimum interval. For example, the new interval cannot be shorter than the old interval (Interval). For a low forgetting index, it is quite common that Optimal interval is shorter than Interval. This is not a reason for worry, but might be an indication that the forgetting index is set too low. The interval after the presented repetition will increase from 2673 days (about 7 years) to 3438 days (over 9 years)
  • Next repetition - date on which the next repetition should take place. Next repetition will be scheduled for 20th December 2013. 2 in the parentheses indicates that there are already two other elements scheduled for review on that day
  • New A-Factor - new value of A-Factor estimated for the displayed item after the just-made repetition. A-Factor was decreased during the presented repetition from 4.672 to 3.283
  • New U-Factor - new value of U-Factor (i.e. the quotient of the new interval and the old interval. U-Factor was changed from 1.564 to 1.286. In other words, the present increase in interval is less than the last increase in the interval back in 1997
  • Expected FI - forgetting index derived from the interval (see the description of the SuperMemo Algorithm). Due to the longer than optimum interval, the expected forgetting index was 11.2% (i.e. slightly more than the requested forgetting index of 10%)
  • Estimated FI - forgetting index derived from the grade (see the description of the SuperMemo Algorithm). From the grade scored in the present repetition, the estimated forgetting index was computed as 16.3%. This value indicates that there would be an increased risk of forgetting the item in the next repetition if the A-Factor had not just been reduced
  • Normalized grade - grade normalized for the optimum interval for the forgetting index equal 10% (see the description of the SuperMemo Algorithm). Here the normalized grade of 3.05 reflects the fact that the grade Pass (3) has been scored after a slightly elongated interval (delay of 1.04). The numbers in the parentheses shows the minimum and maximum values of the normalized grade computed using four different methods
  • R-Factor change - change of the R-Factor corresponding to the current repetition category (the one displayed in parentheses at Repetitions) and A-Factor (displayed at A-Factor among element parameters). See the description of the SuperMemo Algorithm for details. Only grades less than Pass (3) reduce the R-Factor (forgetting pulls the forgetting curve down reducing the interval needed to reach the same forgetting index). In the presented case, the grade 3 increased the relevant entry of the R-Factor matrix slightly (from 2.313 to 2.315). That entry can be found in the RF matrix as RF[11,4.5] where 11 is a rounded value of the repetition category (11.4), and 4.5 is the lower limit of the A-Factor category (for A-Factor 4.672)
  • O-Factor change - change of the O-Factor corresponding to the current repetition category (the one displayed in parentheses at Repetitions) and A-Factor (displayed at A-Factor among element parameters). See the description of the SuperMemo Algorithm for details. For good grades, O-Factors also increase; however, as they come from smoothing R-Factors, these changes are less prominent. In the presented case, the O-Factor has not changed detectably and stayed at the level of 1.4
  • Cases - number of repetition cases used to compute the values of O-Factor and R-Factor corresponding to the current repetition category (the one displayed in parentheses at Repetitions) and A-Factor. Here 801 repetitions have been recorded for repetition category 11 and A-Factors in the range 4.5-4.8

1.3.1