Thresholds
According to Schmeidler, and the vast majority of psychologists studying subliminal phenomena, a threshold is the point at which a stimulus is perceived 50% of the time. For example, if my aural threshold were to be tested, I would be played numerous sounds in varying volumes. To signal that I heard the tone-I perceived the stimulus-I would raise my hand or press a button, whatever the researches had instructed me to do when I heard a sound. By controlling and tracking the frequency of the emitted sound, the researchers are able to find the volume at which I hear a sound-perceive the stimulus-half of the time. This point is my auditory threshold. A working definition of threshold is paramount. It acts as the scientific base for research in this area; a foundation on which all data are defined. For some, this is where the problems with subliminal research begin. xxxx (follow this link for discussion of these issues)

Visual Subliminal Perception

A series of nothings become something
When a person is flashed an image, the brain maintains a type of "footprint or "after image" of the stimulus. That is, despite the stimulus as been shown and then taken away, for a short time that image is still seen by the brain. At first flash the subject may remember the shape of the beginning of a word. The next flash (very close together) the brain creates a footprint of the middle of the word. Continuing in such a way, a subject can piece together these footprints - these 'series of nothings - and form/recognize the word as a whole. (Schmeidler 127)

How visual masking works
The masking of visual perception is generally the taking away of the "footprint" lingering on the subject's brain created by the target stimulus. A nonsense image (a series of letters that spell nothing, a line pattern, etc.) is flashed immediately after the target image. This leaves the footprint of the nonsense image lingering around as opposed to that of the target image. (Epley 7)

Aural Subliminal Perception

Don't steal!!
Many department stores utilize subliminal messages to discourage theft. They use audio messages masked in the "store's music" (many stores play music over their intercom system). Masked audio messages are generally compressed or accelerated to a degree that renders them unintelligible, even if supraliminal. The message is then masked by the playing of, in this case, music. This music is the primary channel - it is the easiest to perceive. The hidden message becomes the secondary channel. (Moore) This tactic may not be futile, "numerous findings indicate not only analysis of secondary channel content at the level of individual words, but short persistence of memory for that content" (Greenwald 5). Perceived or not, there is still the controversy over whether or not it will influence one's behavior.

Difficulties with research

Troubles with thresholds
To begin with, psychologists have essentially massaged the theory of thresholds so that subliminal perception could "exist in a form that can be studied". In the1800's, Herbart and Fechner used the term limen as a distinction between conscious and unconscious. Much of the research conducted during the 1950's and on is based on a distinctive breaking of "threshold" into two parts: sensory threshold and perceptual threshold: stimuli that are "sensed by the body, but not conscious" and stimuli that are "made conscious", respectively. It is the perceptual threshold that is utilized in the realm of subliminal perception research (Erdelyi 3). Thus, what the researchers are studying are stimuli that are above the sensory threshold, but below the perceptual threshold (the point at which a subject can perceive a stimulus 50% of the time). This begs debate for it is sidestepping the entire concept of a dichotomy between conscious and unconscious, the exact concept of subliminal.

Already known from other areas of research is the fact that we do receive information in our lower brain that never makes it to the cortex. The cortex is where things are "made conscious". This is where sensory perception comes in to play. It is estimated that for every 1,000,000 stimuli that pass by the sensory threshold, one stimulus passes through the perceptual threshold (Norr 161). This is how the intricacies of human behavior are explained. Humans simply do too much to be conscious of it all. As McConnell points out, "We are a walking mass of thresholds."

Each person has their own unique set of thresholds. To measure the threshold of each subject prior to conducting the experiments is both time consuming and redundant, but does account for such idiosyncrasies. However, not only do thresholds vary from person to person, but they also vary day by day within one individual. So what I can't hear today I may hear tomorrow and vice versa (McConnell, Moore). With no static threshold, or statistical foundation, research conducted on subliminal perception today becomes invalid tomorrow.

Erdelyi (3) brings up an intriguing point. He asserts that these problems with thresholds are not methodological hindrances, but rather they are conceptual flaws. This reminds the community that indeed this research is not truly "subliminal".

Attention
Attention plays a role on the research of subliminal perception. It is capable of skewing results because the subject is instructed to listen for a sound or to look for a sight. When trying to ascertain if one can be influenced subliminally in day to day life, the person will not be actively looking or listening for the supposed stimuli/message. To counter this, study designs should keep the subject unaware of the target stimuli. (Cohen 55 Shiffrin)