The heightened recall for items at the beginning of a speech (or a list or some other unit of information) is known as primacy. Atkinson and Shiffrin (1968) attributed the primacy effect to the elevated attention and rehearsal directed at the beginning of a speech or list. Such effort helps transfer the information to long-term memory, where the listener can recall it. If you represented the serial position effect as a U-curve, the primacy effect would be the left part of the curve.
The recency effect refers to the recall of information at the end of a speech. Atkinson and Shiffrin theorized that recency signified "output from what they referred to as primary memory in the form of a short-term memory buffer," according to research published by Middle Tennessee University. Because the most recently stated information occupies a position in short-term memory, the listener remembers it more easily. On the same curve, its position would lie on the right, upward-inclined part of the U.
The middle position, between primacy and recency, is, according to the theory, the easiest to forget. Information communicated here escapes both long and short-term memory. Psychologists would term its memorability, in contrast, as "equal recall." In a speech, the middle portion, lacking both a captivating introduction and a culminating conclusion, would be, according to this theory, the least memorable.
The theory (particularly the recency theory) depends on the role of short-term memory. According to psychologists Waugh and Norman, listeners process information sensorily before storing it as short-term memory. From short-term memory, it can enter long-term memory for longer recall. Delaying recall wipes out the recency effect, according to the University of Indiana.