The “forgetting curve” is a term psychologists use when referring to the mental process by which information once known is gradually eroded. There are a number of methods for measuring the forgetting curve, also known as the “curve of forgetting,” but generally involves exposing test subjects to information and then tracking the period of time it takes for them to forget elements of that information. Part of the testing involves refreshing the memories of the test subjects are certain intervals to then further gauge how much of the information is retained over specified periods of time. A German psychologist named Hermann Ebbinghaus (1850-1909) is considered the father of the concept of a forgetting curve, using himself as his initial test subject and plotting on a graph the decline in how much of a series of three-letter words he was able to retain over periods of time. Beginning his graph with 100 percent retention of newly-reviewed material, he would then track the rate at which that material was forgotten over time. In general, over the course of one year, most information not reviewed intermittingly was lost or forgotten.
Ebbinghaus’ research and experimentation also involved manipulating the process to determine what intervening variables could influence the outcome, for example, altering the type of material used to better reflect the knowledge of the individual test subject and presenting the information in various ways to best identify presentations of information that will be retained for longer periods of time. Ebbinghaus’ research has been instrumental in developing methodologies for use in the classroom to help students learn.
In a field of psychology pioneered by Hermann Ebbinghaus, a German philosopher whose Ph.D. dissertation was Philosophy of the Unconscious, the forgetting curve is the measure of the rate at which information is forgotten. This groundbreaking research, which psychologist William James called "heroic," is important because it proved for the first time that higher order mental or psychological processes can be studied through empirical experimentation, a concept that was previously thought impossible. Ebbinghaus's research results were additionally groundbreaking because he effectively measured the (1) exponential curve that charts the rate at which new nonassociative information is forgotten, the (2) similarly exponential rate at which new nonassociative information is learned, the (3) the serial position effect of recency and primacy whereby the most recent and the first in order are remembered best, and the (4) savings in learning effect consequent of information retained in the subconscious though not consciously accessible.
The research method Ebbinghaus used was an empirical one not a theoretic philosophical one. He developed a collection of 2300 tri-part consonant-vowel-consonant syllables that had no associations to previously learned information. This means that the tri-part syllables GET and PED, for example, were not acceptable for Ebbinghaus's research because they both had associations with prior learning: "GET" is an actual word, while "PED" is the beginning syllable of several words, like pedestrian and pedagogic. Only new nonassociative syllables were acceptable, like LIJ or KOJ, as Ebbinghaus was interested in pure, nonassociative learning. It is regrettable that Ebbinghaus did not save his list of 2300 syllables, which later came to be known as nonsense syllables, for future researchers so we might at least have examples of with what Ebbinghaus himself worked.
Ebbinghaus used these syllables for memorization--with himself as his only subject rendering his results not generalizable--and for recall. He carefully measured (1) the times spent on memorization and (2) the results of recall at timed intervals. One innovation in his study was designed to control the variable of the mental, psychological and physical condition of the subject, himself. Preoccupation, distractions, and fatigue etc all affect the rate of learning, memory and recall. Ebbinghaus developed a method that controlled this variable: he recited his randomly drawn lists of syllables not for a specified time nor to a specified number of repetitions but until he had attained perfect recall. This insured that the level of memorization was the same in trial after trial involving set after set of tri-part syllables. This method is now called learning to criterion based on a preset a priori level of achievement.
Ebbinghaus's results showed that both learning and forgetting progress at predictable exponential rates. This means that across time, as may be shown in a graph and studied statistically, the rate of progress or change increases or decreases as time progresses: the more time, the more change. Compare this to linear rate of change in which, at any time interval, the rate of change is the same as at any other time interval. Specifically, Ebbinghaus's results show that the greatest rate of forgetting--the sharpest decline in the loss of information--occurs in the first 20 minutes after learning (for Ebbinghaus, this meant in the first 20 minutes after reaching perfect recall). More than half of the information is forgotten by the end of the first hour. The forgetting curve levels off, with a scant amount transferred to long-term memory, by the end of the first day after a vast majority of the information is lost. Ebbinghaus's results also showed that "savings" (the retention of information in the subconscious) lessen the forgetting curve with each subsequent trial at memorizing the same information. Thus because of savings, with each trial, (1) less is forgotten and the curve is less steep and (2) learning is best accomplished over time with multiple sessions.
When learning new information that has associative value--that is information associated with previous learning--the forgetting curve is less steep with less forgotten during predictable intervals. This is because, though there is still a forgetting curve, new information is associated with or connected with old information thus reducing the exponential components of the forgetting curve. Consequently, this associative connectivity of new and old information accounts for the decrease in the forgetting curve in repeated memorization trials with new nonassociative information: as the new information transfers to long-term memory, it takes on characteristics of previously known, or old, information and permits greater and greater associative connectivity at each subsequent trial.
First Image: Ebbinghaus Forgetting Curve
Second Image: Illustration of Exponential Decline as in the Forgetting Curve