Additional Research Topics
In the research literature, you will find studies that deal with
almost every behavior imaginable. Below are some common and not so
common research topics with which you can become involved rather
easily.
Belief in Astrology
Astrological horoscopes and personality descriptions contain Barnum statements such as ``You are generally a happy person, although you sometimes become angry.'' A person's reactions to such descriptions tells us not only about the popularity of astrology but also about general belief systems. For example, Glick, Gottesman, and Jolton (1989) studied ``believers'' and ``skeptics'' to determine how they deal with positive and negative descriptions in horoscopes. The researchers proposed that both types of participants would be impressed with the accuracy of positive descriptions (e.g., ``You are intelligent''), presumably because such descriptions affirm positive self-perceptions. But they also suggested that only believers in astrology would accept negative descriptions (e.g., ``You are indecisive''), because their faith in astrology overrides their self-perceptions. In addition, the researchers tested whether a description not attributed to astrology would be accepted as more accurate, and whether experiencing positive or negative descriptions would alter participants' belief in astrology.
What type of design would you use to study these factors? Testing any of these factors involves first the quasi-independent variable of creating conditions by selecting participants who are astrology believers or skeptics. Additional factors then include favorable versus unfavorable personality descriptions, and/or whether or not the descriptions are attributed to astrology. The dependent variable would be participants' rating of the accuracy of the descriptions. In their study, Glick et al. (1989) manipulated these factors as between-subjects factors, so that they could also examine the influence of a particular type of experience on participants' attitude toward astrology. They tested this influence with a pretest-posttest design, measuring participants' attitudes toward astrology before and then after they participated in the above conditions.
The researchers found that (1) skeptics accept positive personality descriptions as more accurate than negative ones, but believers accept a negative description as being equal in accuracy to a positive one, (2) believers and skeptics alike rated astrological descriptions as more accurate than nonastrological descriptions, and (3) skeptics, more so than believers, became more positive in their feelings toward astrology after receiving favorable horoscopes. Glick et al. proposed that skeptics are more open-minded, so they are more likely to change their opinion about astrology after receiving a positive description. Believers, however, seem to ignore the contradictions from a negative description, thus suggesting that believers and skeptics differ in terms of how they test these descriptions and evaluate the evidence that may support them.
What suggestions for additional research can you make? First, it would be interesting to examine how believers and skeptics test astrological predictions. For example, you might present them with a horoscope that is ultimately confirmed or disconfirmed by what ``coincidentally'' happens later in an experimental setting. Then you could measure how the outcome of a prediction is evaluated by skeptics and believers, and how this evaluation influences their attitudes toward astrology.
Second, when believers are confronted with a negative statement
that disconfirms their self-perceptions, they continue to maintain
their original opinion. The implication is that believers in
astrology do not apply the correct logic for testing hypotheses.
Thus, you could test for differences in logic and hypothesis testing
between skeptics and believers.
Serial-Position Effects
A highly reliable finding in the study of memory is that when participants recall a list of words, they tend to recall the first few and the last few items in the list best. Because recall changes as a function of an item's serial position in the list, this effect is called the serial position effect. Often in these studies, a list is spoken to participants which consists of a random string of single digits, and, for reliability, each participant is tested with a number of lists. To see this effect, we graph the participants' recall scores for the words in a list as a function of where in the list the words occurred. The graph produces a serial-position curve, as shown in Figure 23.2.
Note that the higher recall of the first items in the list is called the ``primacy effect'' and the higher recall of the final items is called the ``recency effect.''
To study the memory processes that produce these effects, researchers have employed numerous techniques that alter recall of the list. From the impact of a particular manipulation, they infer a characteristic of memory. For example, the final few items in a list are most recently entered into memory, so they may be in some sense ``fresher.'' Some researchers (e.g., Crowder, 1982) have proposed that these items are better recalled because there is an ``echo'' in memory of the most recent spoken words. The final word in the list is recalled best because no words come after it that might otherwise interfere with its echo.
How would you test this explanation?
Researchers have tested this proposal by speaking an additional word at the end of the list, called a ``stimulus suffix.'' When the lists consist of digits, the suffix for every list is usually the digit zero. So as not to increase the effort that participants must apply to remembering the list, they are told not to remember the zero; it is merely the signal to recall the list, and they are to otherwise ``ignore'' it. In control conditions, a tone is the recall signal. Because the tone is so different from the list of words, it should not produce interference, so recall of the list in the control group should exhibit the typical high recency. Conversely, because the zero is heard, it should reduce the freshness of the final list items and fill the echo, so recall of the list in the experimental group should exhibit reduced recency relative to the tone condition. In fact, it does, and the lowered recency is called the ``suffix effect.'' (See Turner et al. [1987] for an overview.)
The reason that the suffix disrupts recency is not known. Originally, its effect was thought to be limited to spoken words only. Yet suffix effects have also been found when the zero is silently mouthed by the experimenter or by the participant, and among deaf participants when the list and suffix is presented using sign language. In addition, suffix effects have been shown to occur when spoken words in the list have different vowels (gap, got, gut) but not when they have different consonants (gap, tap, map). Thus, the suffix effect is not restricted to auditory stimuli, yet it does not always work with auditory stimuli. This inconsistency makes the notion of an ``echo'' very suspect.
Although there have been many explanations of this effect, one that has received virtually no attention concerns the assumption that participants can in fact ``ignore'' the zero at the end of the list. Participants are presumably equally able to ignore a tone and ``zero,'' and both readily signal that the list is over. Then the suffix has its effect simply because in some way it fills the echo, decreasing the freshness of the final items in the list. But how do participants know that the word zero is the one to ignore unless they first pay attention to it? They must identify this word to recognize that it is not part of the list to be remembered. A tone, by contrast, can be ignored instantly, because it is so obviously different from the preceding list of words. Can (or do) participants ignore the zero? If not, is the zero at first treated as part of the list, thereby confusing participants? Does a list followed by a zero require greater attention than a list followed by a tone, and does this greater attention come at the expense of recalling the list?
How would you study whether participants do ignore the suffix? You might measure the reaction time of participants to determine whether there are differences in their ability to recognize the end of a list when it is signaled by a tone or zero. If the zero takes longer to recognize, then there is something about it, in terms of the attention it requires, that is different from the control condition. If so, then a confounding has occurred, because the tone and the zero are not equivalent signals for the end of the list.
Alternatively, you might first identify some words that are easy
to recognize and some that are difficult to recognize. Then, using
these words in place of the zero, you can see whether they alter the
recency effect. Or you might give participants a cue to forewarn them
that the zero or tone is about to occur. For example, you might ring
a bell just prior to the zero. If reading the list aloud, you could
change your tone of voice and inflection, or make a movement, to
signal that the zero is about to occur. If such cues eliminate the
suffix effect, then you have evidence that the zero normally reduces
the recency effect because of the attention it demands or the
confusion that it produces.
Attractiveness and Height
Evidence suggests that, in our society, taller men are perceived as more honest, competent, and attractive. For example, men described as tall when advertising in the ``Personals'' column of a newspaper tend to receive more responses (Lynn & Shurgot, 1984). Accordingly, Shepard and Strathman (1989) investigated whether women prefer to date taller men and whether they consider taller men more attractive. They also investigated whether men prefer to date shorter women and whether they consider shorter women more attractive.
How would you study the relationship between height and attraction? Shepard and Strathman conducted a correlational study, presenting males and females with a questionnaire that asked for the number and height of their recent dates as well as for a rating of each date's attractiveness. Participants were also asked whether they preferred to date a person who was shorter than, taller than, or the same height as themselves. The researcher's found that females reported that they dated tall males more frequently than short or medium males. They also reported a preference for taller men, but they did not rate their taller dates as more attractive. Interestingly, short and medium-height males reported that they went out on dates just as often as did tall males. Overall, the males preferred shorter female dates, rated them as more attractive, and dated them more frequently.
What problem do these self-reports present?
These participants' self-reports may have been unreliable for several reasons: (1) The subjects might have inaccurately estimated a date's height, especially after the fact; (2) the date's personality and compatibility might have influenced participants' perception of the date's attractiveness; and (3) a ``bad'' date might have biased participants so that they remembered their dates as shorter or taller than they actually were. Further, as Shepard and Strathman noted, the females reported fewer dates with short or medium-height men, but the males, regardless of their height, reported the same frequency of dates. Since the shorter males were dating somebody, the self-reports of either the males or the females must have been in error.
What suggestions for additional research can you make?
Rather than relying on self-reports about participants' past dating experiences, Shepard and Strathman conducted an additional experiment in which they manipulated height and measured attractiveness. Here, they presented participants with a photograph of the upper bodies of a male and female facing each other. In the photograph, the male was either 5 inches taller, the same height, or 5 inches shorter than the female. Females rated the man as more attractive when he was taller. Males, however, did not rate females differently depending on her height.
You might also determine why the females' reports of seldom dating short men does not correspond to the high frequency of dating reported by short men. Perhaps the males or females are erroneous in their reports because of pressure to give the socially desirable response. If so, the number of erroneous reports should increase when the researcher strongly implies such demand characteristics through instructions or the task. Also, you might investigate whether the contradictory results occur because of a definitional problem, having to do with the term taller. Stereotypically, most people would agree that a man whose height is 7 feet is ``tall,'' and that one who is 5 feet is ``short.'' However, taller can be a relative personal term, because for a woman whose height is 4 feet 6 inches, a man with height of 5 feet is ``taller.'' To what extent do males and females use a stereotypic or personal definition when selecting or describing dates?
There is an ecological issue here as well: Women are, on
average, shorter than men. The opportunities for dates are greater,
then, if women accept taller men and men accept shorter women, so
their ``preferences'' may simply reflect the facts of life. Does the
fact that most men are taller than most women lead to a stereotype of
the ideal date? As we saw when we discussed Horner (1972), people
tend to anticipate negative consequences when violating a sex-role
stereotype. To what extent do the above studies measure stereotyped
responses instead of actual attraction? And, finally, is it
appropriate to conclude that being taller always makes a male more
attractive to a woman? Is there a point where taller is not better
but becomes ``too tall''?
Influence of Color
Research from environmental psychology suggests that the color of walls, furniture, or floors has an influence on various aspects of behavior. For example, ``warm'' colors (those close to red) are often believed to be arousing, to increase physical performance, and to improve mood. And ``cool'' colors (those close to blue) are believed to be soothing, to lower performance, and to have a dulling effect on mood.
How would you design a study to test these beliefs?
One obvious way to manipulate the color in the environment is to test participants in different rooms that are painted a different color. In each room, you can have participants perform a physical task and/or provide responses that indicate their mood. For example, Kwallek, Lewis, and Robbins (1988) asked participants to type business forms for 20 minutes after placing them in either a red or blue ``office,'' and then had them complete a questionnaire describing their anxiety, mood, and general arousal. After resting in another room, participants either returned to the same colored office or were switched to the different colored office, performed additional typing, and completed another questionnaire. As it turns out, the participants who were moved to the different colored office made significantly more typing errors. And the interaction showed that more errors occurred when participants moved from the blue to the red office than when they moved from red to blue. The mood data indicated that participants remaining in the red office showed greater anxiety and stress, participants remaining in the blue office showed greater depression, and participants who were moved to different colored offices showed the greatest level of general arousal.
Because moving the participants into a different colored room influences their behavior, a problem with this study is that placing participants in the first office constitutes changing rooms in an uncontrolled manner. As an alternative approach, you might include a control condition that allows you to ``reverse'' the influence of any previous colored room.
How would you design this alternative study?
Hamid and Newport (1989) studied how the color of a room influences behavior using an ABACAB reversal design. The control or baseline---condition A---was a gray room that presumably neutralized the influence of other colors. A pink room was condition B, and a blue room was condition C. The researchers measured hand-strength and mood in six young children after they had experienced each colored room. Greater physical strength and more positive mood was found in the pink conditions.
What suggestions for additional research can you make?
If there is an influence of color on arousal and performance, then
it should extend to a variety of colored objects and situations.
Would similar effects on typing errors occur depending on the color
of the paper in a typewriter or the color of a computer display
screen? In the previous chapter we saw the influence of black
uniforms on aggression, but the present research suggests that other
colors of clothing may also affect behavior. Certainly the idea
behind ``dressing for success'' is that a worker's or job applicant's
clothing style influences how that person is perceived and judged by
supervisors and co-workers (e.g., Forsythe, 1990). Does wearing
reddish or bluish clothes also influence perceptions? We also
discussed how participants misattribute their arousal due to fear as
being due to sexual attraction. Is it possible that arousal from wall
color or the color of clothing could be misattributed as sexual
attraction? As an applied topic, the colors in business environments
may be important for maximizing worker productivity. Similarly, given
that maximum arousal is sought in athletic events, does wall color
influence, for example, weight-training success? And, finally, do
certain colors play a role when we become overaroused, as discussed
below?
Self-Consciousness and "Choking Under Pressure"
An unusual behavior to study is the phenomenon of ``choking under pressure.'' Baumeister (1984) proposed that inferior performance (``choking'') occurs when we feel so pressured to perform well that we focus too much attention on the process of performing a task and not enough attention on the outcome of the task. Thus, in a self-fulfilling prophecy, the more we worry that things are going badly, the worse they go. Baumeister also proposed that personality characteristics play a role, such that, for example, a person who is more self-conscious should exhibit greater ``choking.''
How would you design a study to test these proposals?
In an experimental setting, you could create a task for participants in which you manipulate the amount of pressure they feel. The task should be a simple one in which errors are easily measured. For example, you can have participants quickly trace a pencil maze, solve simple math problems, or perform a reaction-time task in which they press one of several buttons to make a correct response. Consider Heaton and Sigall (1991) who first identified participants along the quasi-independent variable of high or low self-consciousness. For experimental realism, they had participants in each condition form a ``team.'' Then they manipulated the pressure situation through the additional factors of: (1) indicating that the participant's team was behind or ahead, and (2) having participants perform when alone, when watched by their team, or when watched by the opposing team. Choking was measured by the time it took to place variously shaped pegs in their corresponding holes. Participants low in self-consciousness choked depending upon the audience characteristics, while those high in self-consciousness choked when their team was behind.
What suggestions for additional research can you make? Apparently, people differing in self-consciousness perceive the source of pressure differently. The results for participants with high self-consciousness suggest that they choked because of competitive pressure, so you might manipulate the situation for them in terms of the amount of competition involved. The results for participants with low self-consciousness suggest that they choked because of their need for social approval, so for them you might manipulate the social setting. Also consider the possibility of replicating the above relationship between choking and pressure in a field setting: Perhaps a ``stranger'' could watch someone playing a game in a video arcade. As extensions of this research, the above results suggest that other personality traits, in terms of whether a person is task or socially oriented, may play a role in choking. Thus, for example, you might correlate participants' scores on these traits with how much they choke or, as above, examine how these traits interact with different conditions that promote choking. Alternatively, given that high pressure is a form of arousal, you might ask whether the color of the room in which a participant is being tested influences choking. Finally, you might ask whether a ``lucky charm,'' such as a lucky shirt, is considered lucky because a participant did not choke when wearing it. (And what color is it?)