In the study of perceptual abilities of infants, a number of techniques are used to determine infants' responses to various stimuli. Because they cannot verbalize or fill out questionnaires, indirect techniques of naturalistic observation are used as the primary means of determining what infants can see, hear, feel, and so forth. Each of these methods compares an infant's state prior to the introduction of a stimulus with its state during or immediately following the stimulus. The difference between the two measures provides the researcher with an indication of the level and duration of the response to the stimulus. For example, if a uniformly moving pattern of some sort is passed across the visual field of a neonate (newborn), repetitive following movements of the eye occur. The occurrence of these eye movements provides evidence that the moving pattern is perceived at some level by the newborn. Similarly, changes in the infant's general level of motor activity —turning the head, blinking the eyes, crying, and so forth — have been used by researchers as visual indicators of the infant's perceptual abilities.

Such techniques, however, have limitations. First, the observation may be unreliable in that two or more observers may not agree that the particular response occurred, or to what degree it occurred. Second, responses are difficult to quantify. Often the rapid and diffuse movements of the infant make it difficult to get an accurate record of the number of responses. The third, and most potent, limitation is that it is not possible to be certain that the infant's response was due to the stimulus presented or to a change from no stimulus to a stimulus. The infant may be responding to aspects of the stimulus different than those identified by the investigator. Therefore, when observational assessment is used as a technique for studying infant perceptual abilities, care must be taken not to overgeneralize from the data or to rely on one or two studies as conclusive evidence of a particular perceptual ability of the infant.

Observational assessment techniques have become much more sophisticated, reducing the limitations just presented. Film analysis of the infant's responses, heart and respiration rate monitors, and nonnutritive sucking devices are used as effective tools in understanding infant perception. Film analysis permits researchers to carefully study the infant's responses over and over and in slow motion. Precise measurements can be made of the length and frequency of the infant's attention between two stimuli. Heart and respiration monitors provide the investigator with the number of heartbeats or breaths taken when a new stimulus is presented. Numerical increases are used as quantifiable indicators of heightened interest in the new stimulus. Increases in nonnutritive sucking were first used as an assessment measure by researchers in 1969. They devised an apparatus that connected a baby's pacifier to a counting device. As stimuli were presented, changes in the infant's sucking behavior were recorded. Increases in the number of sucks were used as an indicator of the infant's attention to or preference for a given visual display.

Two additional techniques of studying infant perception have come into vogue. The first is the habituation-dishabituation technique, in which a single stimulus is presented repeatedly to the infant until there is a measurable decline (habituation) in whatever attending behavior is being observed. At that point a new stimulus is presented, and any recovery (dishabituation) in responsiveness is recorded. If the infant fails to dishabituate and continues to show habituation with the new stimulus, it is assumed that the baby is unable to perceive the new stimulus as different. The habituation-dishabituation paradigm has been used most extensively with studies of auditory and olfactory perception in infants. The second technique relies on evoked potentials, which are electrical brain responses that may be related to a particular stimulus because of where they originate. Changes in the electrical pattern of the brain indicate that the stimulus is getting through to the infant's central nervous system and eliciting some form of response.

Each of the preceding techniques provides the researcher with evidence that the infant can detect or discriminate between stimuli. With these sophisticated observational assessment and electro-physiological measures, we know that the neonate of only a few days is far more perceptive than previously suspected. However, these measures are only "indirect" indicators of the infant's perceptual abilities.

在婴儿感知能力的研究中，许多技术被应用于确定婴儿对不同刺激的反应。由于他们（婴儿）无法用言语表达或者填写问卷，所以自然观察的间接技术被应用成主要的方式去确定婴儿的视觉、听觉和感觉,等等。每种方法都是将刺激引入前婴儿的状态和刺激中或刺激之后婴儿产生的反应作对比。两种标准的不同性给研究人员提供了对刺激反应程度和反应持续时间。比如说，如果一个移动物体经过婴儿（新生儿）的视野，他们就会反复转动眼球。。这种眼球转动的现象可以说明移动物体在一定程度上引起了新生儿的注意。同样，新生儿一般身体活动的改变——比如摇头、眨眼、哭闹或者其他，都可以为研究人员提供婴儿感知能力研究的直观参考。 然而，这些技术是有局限性的。第一，两个甚至更多的观察者也许不会一致认可特殊反应的发生或者反应的程度，因此这种观察可能并不可靠。第二，反应很难量化，婴儿的很多反应快速且零散，研究人员很难准确记录。第三点也是最重要的一点，我们很难确定婴儿的反应是由现存的刺激或者是因为从无刺激到有刺激的改变。可能引起婴儿反应的并非是研究者认为的刺激方面。因此，当观测式评估被用作研究婴儿感知能力的技术时，必须注意不要从资料中过度归纳出结论，或（不要）将一个或两个研究作为婴儿一项特殊感知能力的决定性证据。 观察评估技术变得更加复杂，上述局限性也在降低。婴儿反应的影片分析、心脏和呼吸速率监视器以及无营养允吸设备都是理解婴儿感知能力的非常有效的工具。影片分析允许观察者反复或者慢镜头播放来仔细研究婴儿的反应。通过观察婴儿在两次刺激之间注意力的长度和频率可以完成精准的测定。而心脏和呼吸频率监视器可以让观察者了解新刺激产生时婴儿的心跳次数和呼吸频率。数值增加被当做是新刺激中提升兴趣的可量化指标。1969年，研究者们首次将无营养吮吸设备动作次数的增加作为评估标准。他们设计出了一种连接婴儿无营养允吸设备和计数器的设备。只要出现刺激，这种设备就会记录下婴儿的吮吸习惯。吮吸次数增加就说明某种视觉展示引起了婴儿的注意或者婴儿偏好此种视觉展示。 另外两种研究婴儿感知的技术也逐渐开始流行。第一种是习惯与非习惯技术：给婴儿反复展示一种单一的刺激，直到观察到婴儿对这一刺激形成习惯并出现可测量的减弱（习惯性）。接着当出现一种新的刺激时，任何反应的恢复也都会被记录下来（非习惯性）。如果婴儿对新刺激没有不习惯而是继续表现出习惯性，那么我们就可以推测婴儿没有办法识别出新的刺激有什么不同。这种习惯与非习惯的技术在婴儿的听觉与嗅觉感知能力的研究上广泛运用。第二种技术基于诱发电位，大脑的弱电反应可能因其起源而与特定刺激有关系。脑电图发生变化意味着刺激通过婴儿的中枢神经系统，引起某种形式的反应。 上述每种技术都可以给研究者提供证据，证明婴儿能够探知或辨别刺激。通过这些复杂的观察评估和电生理学的测量，我们知道刚出生几天的新生儿的感知能力超乎我们的想象。然而，这些测量也都仅仅是婴儿感知能力的“间接”指标。