In this final chapter you will be provided with fuller answers to a couple of essay or exam questions, and one laboratory report. The purpose is to guide you through the essay writing processes and to highlight some important features of writing good essays. The lab report will illustrate how to construct each subsection of a report. Question 1 concerns the material covered in Chapter 3, Section 1. Question 2 provides new material (research on recovered memories and false memories) but shows how material from Chapter 5 can be drawn on. Question 3 is a research report on a hypothetical practical that brings together material from Chapter 4 and Chapter 5.

Question 1: Discuss two theories of focused attention

[For your first few sentences, state what the problem is and what behaviour needs to be explained. Give an everyday example, if relevant.] The study of focused attention is concerned with how we are able to selectively attend to some information in the environment, ignoring other information. An example of this problem was described by Cherry (1953) and is known as the ‘cocktail party effect'. When we are at a party, we are able to focus on what someone is saying to us, despite a lot of background noise (other people talking and loud music). We can effectively ignore this background noise. However, should someone, other than the person we are talking to, mention our name then we are likely to hear it immediately. This effect highlights the difficult question of how we can ignore something, yet still be attending to it. [Tell your reader which approaches your essay will focus on. State what the main theoretical approaches are in attempting to explain the behaviour.]

Several theories have been proposed that attempt to explain this and other attentional phenomena, and this essay will discuss two theories of focused auditory attention, namely Treisman's attenuation model (Treisman, 1964) and the late selection theory of Deutsch and Deutsch (1963). [Take a historical approach, if there is much research on the topic (e.g. as in focused auditory attention); otherwise take each theory in turn beginning with the weakest.]

Both the attenuation model and the late selection theory use variations of the dichotic listening task in conjunction with the shadowing task. Participants are presented with two auditory messages, one to each ear via a set of headphones. They are required to attend to one message while ignoring the other message. To be sure that participants are attending to the message they are asked to repeat aloud the message they hear (the task known as ‘shadowing'). Typically, the messages presented are of spoken words or sentences.

Researchers wish to know how much and what type of information the participant can hear from the non-attended message (the ‘unshadowed' message). Early studies using this method suggested that little information is attended to in the unshadowed message. These led theorists such as Broadbent (1958) to suggest that in focused auditory attention, we filter out the ignored message so that we can focus on the 133 relevant message. In the sensory channels, too much information creates a bottleneck and this problem is solved through the process of filtering. However, other findings suggested that some information, such as one's own name, the gender of the speaker and other features could be detected in the unshadowed message. This led Treisman to develop her attenuation model of attention (Treisman, 1964). [Describe the approaches with some evaluative statements. Describe the theory and any evidence that supports it. Next, describe any research that does not support the theory or any behavioural observations that the theory cannot account for.]

According to Treisman, the unattended message is not entirely filtered out but just ‘tuned' down. Physical (shallow) features of the stimulus are processed first, followed by stimulus meaning (deep features). The more processing capacity that is available, the deeper the unattended message can be processed. This fits in with Cherry's (1953) observation that the physical characteristics of unattended inputs, such as the sex of the speaker, can be noticed more often than their meaning, since usually there is sufficient processing capacity for this. Research that supports the model comes from Treisman (1960) in which bilingual speakers of French and English heard an English message in one ear and a French translation in the other ear. When the messages were presented at about the same time, participants recognised that the two messages had the same meaning. However, they did not recognise this when there was a time delay between the two messages. In another study, Treisman (1964), it was found that the participants' ability to shadow the attended message was affected by the degree of similarity between the attended and unattended ear. The more similar the passages were (e.g. different passages from the same novel), the harder it was to shadow the attended message. These studies suggest that the unattended message is processed for meaning to some extent.

There may be problems with the Treisman model. However, the model of Deutsch and Deutsch (1963) needs to be discussed next before the research evidence can be evaluated. [Next, describe a second theory, saying in what ways it is similar to and different from the first theory. Provide evidence that supports it and evidence or observations that do not support it. Do the same for a third theory, if appropriate.] Deutsch and Deutsch (1963) proposed a model in which all stimuli are fully analysed, with the most importantmessage determining the response. While this theory has the notion of a bottleneck as in Treisman's model, it places its location closer to the response end of the system. It is therefore known as a late-selection theory. As the decision is made late, it assumes that most processing is parallel up to that point.

The model predicts that since selection is made much later than in the Treisman model, then some information in the unattended channel may be processed at least nonconsciously. The study by Von Wright et al. (1975) supports this prediction, since in that study a physiological response was recorded during the presentation of significant stimuli to the participant in the unattended channel. The main difference between this model and that of Treisman's is essentially that in the former all stimuli are processed for meaning, while in the latter the available resources determine the amount and depth of processing. [Provide a critical evaluation of the approaches. Finally, highlight the strength and weaknesses of each theory. Then provide an overall evaluation of which theory accounts for the behaviour, outlining its main strengths.]

Determining which model accounts for the available research evidence better may be difficult. For example, in Treisman and Geffen (1967), the participants shadowed one of two auditory messages, and were told to tap whenever they detected a target word in either message. Attenuation theory predicts that participants will be less able to detect the target in the unattended channel since most attention is directed at the attended channel and the unattended channel is attenuated. However, the Deutsch and Deutsch model predicts that detection of the target should be the same regardless of whether it is presented in the attended or unattended channel. Results indicated that detection rates were significantly better in the attended channel (87 per cent versus only 8 per cent in the unattended channel), therefore supporting Treisman's model. However, Deutsch and Deutsch argued that participants had to shadow the message and be vigilant for targets in the attended channel, but only had to listen out for targets in the unattended channel, and therefore the processing priorities were not matched. In response to this criticism, Treisman and Riley (1969) replicated the study but required that participants stop shadowing when they heard the target. In this way the demands made on both channels were more equal. Once again, though, significantly more targets were detected in the attended channel than in the unattended channel, providing support for Treisman's model.

The study by Johnstone and Wilson (1980) provides some evidence that both models may be partly correct. Pairs of words were presented and the task was to identify target words (e.g. sock) that were members of a given category (e.g. articles of clothing). Each target word was presented with another word (e.g. smelly or punches) that would bias its meaning. When participants did not know which ear the target would be presented in, target detection was improved by appropriate other words that biased its meaning. Therefore, participants were processing both channels for meaning. However, when participants were told to attend to the words in one ear only, target detection was unaffected by the other word that was supposed to bias its meaning. This suggests that non-target words presented in the unattended ear were not processed for meaning. These findings indicate that the amount and depth of processing of non-attended channels is more than is necessary to perform the main task. In sum, it suggests that more information in the unattended channel is processed than is suggested by Treisman, but less than is suggested by Deutsch and Deutsch.

In conclusion, neither of the models suggested by Treisman (1964) and Deutsch and Deutsch (1963) may represent the final word on focused auditory attention. These models mainly differ in terms of the point at which selection is made, rather than on what basis selection is made. The work of Johnston and colleagues (Johnstone and Heinz, 1978; Johnstone and Wilson, 1980) may provide a more plausible account of the process of focused attention. It offers a new, but subtle, account of the way in which psychologists think about focused attention, which is that attention may ‘switch' between signals, rather than actively or prematurely selecting one signal and ignoring another. In addition, recent research on divided attention has put doubt on the notion that there is an attention bottleneck in the first place. Attention may be dependent upon the particular task being carried out and the amount of skilled behaviour involved. As such, attention seems to be more flexible than proposed by the two models discussed here. [Ideally, the answer would include descriptions of a handful of additional studies, or more details of the models presented – see Chapter 3, Section 1.]

Question 2: Is it possible to remember something that did not happen?

Evaluate some of the evidence that has been gathered to answer this question

[This question concerns an ongoing debate about whether ‘false memories' can exist and some of the material covered in Chapter 5 will be useful in answering it]. A false memory can be defined as a ‘memory' of something that did not happen (Neath, 1998). Although there is substantial evidence in memory research of the inaccuracies of memory, the current debate about the existence of false memory was instigated by the controversy surrounding recovered memories. [Note the essay begins by applying and defining a psychological term (false memory) that is used for the phenomenon of remembering something that did not happen.]

There are a number of famous cases where the recovery of long lost memories has led to a conviction (Loftus and Ketchum, 1994). Many of these cases involved adults who recovered lost memories of being sexually abused as children. These memories may reflect events which actually occurred, but they are often difficult to verify by others. Why they have been forgotten for many years is often explained by the psychoanalytic construct of repression. Painful memories of being abused by parents or relatives get pushed into the unconscious because they are difficult to deal with. At some point in the future, the memory may re emerge, cued either by something in the person's environment or by a therapist. [The next paragraph lays out a little bit of history behind the research and outlines one claim, that recovered memories are of events that actually happened.]

Some memory researchers, familiar with the research on the inaccuracies of memory, suggest that claims of recovered memories should not be accepted without an element of doubt. For example, Loftus (1993) points out that in an effort to help their clients, therapists may unintentionally mislead the client into believing that a traumatic event must have occurred in the client's childhood. Furthermore, although people who report recovered memories also report being very confident in them, there is much evidence that confidence in the accuracy of a memory and actual accuracy can often be weakly correlated. Several studies have shown that people can be misled into recalling something when they are first provided with misleading information . [This paragraph presents an alternative claim that recovered memories might not be accurate, or worse, they may be false memories. The next paragraph provides some evidence for this alternative claim concerning whether it is possible to remember something that did not happen. I have lifted the phrase from the question here to remind you that in a topic such as this it is all too easy to stray from the question.]

Strong evidence that misleading information can yield a false memory comes from Loftus (1993) where people were misled into believing that a traumatic event had occurred in their childhood. In one case, Chris was asked to recall events from his childhood after reading a log of his childhood compiled by his brother. The brother was asked to include a event that did not occur, that of being lost in a shopping mall. Not only did Chris (falsely) recall being lost in a shopping mall, he also reported the event in detail, citing how fearful he had felt and also the physical features of the stranger who came to his aid. Chris reported the event as though it was a real memory, when it was an experimental manipulation and hence a false memory. Evidence like this raises the possibility that recovered memories may be wrong or may have been implanted in some way. [You could add description of eyewitness testimony from Chapter 5, Memory in the long term.]

Earlier research on reality monitoring provides an interesting way of showing how memory for an event can become confused. Participants were presented with a list of words, some of which were read aloud by the experimenter and some of which the participant was asked to imagine being read aloud by the experimenter. In this case, the source of each word was either externally or internally generated. A reasonable assumption is that externally generated information is associated with a broad set of attributes linked to sensory properties, while for internally generated information this is less so. One should expect, therefore, that participants could easily discriminate between external events and internal events when required to identify the source of each word in the list. However, when participants were instructed to do this there were a significant number of errors or confusions (Johnson et al., 1988). This research suggests that discriminating between real and imaginary events can sometimes be difficult. [This paragraph introduced research from another type of study and shows a breadth of knowledge and also adds weight to the argument.] Monitoring the source of an event may involve an inferential process: qualities of a memory are used to infer its source, and richer qualities are associated with external rather than internal sources. There is much evidence from research in episodic memory to suggest that when individuals attempt to date an event, they also make inferences when reaching a decision. [You could add a discussion of episodic memory here – see Chapter 5: Memory in the long term.]

Other laboratory research has shown that people can create a false memory for words they have not been shown. Roediger and McDermott (1995) presented a long list of words (e.g. table, sit, legs, butter, sandwich). Each word was associated with a target word, such as chair or bread, which was not shown. After the list was presented, participants had to rate how confident they were that each word from a second list had been shown earlier. The second list contained words from the original list but also non-presented words.

There were two types of non-presented words, target words (such as chair and bread) and non target words that were semantically unrelated (e.g. boy). They found that 86 per cent of presented words were rated as being shown in the original list, and remarkably 84 per cent of non-presented target words were rated as being shown in the original list. Participants had then created false memories of words that had not been shown. These intrusions have been found with a variety of stimuli (Robinson and Roediger, 1997; Brainerd and Reyna, 1998). One could conclude that similar intrusions occur with childhood memories: the details recalled may be similar to actual experiences but never occurred. [It is important to give details of important studies like this one. Note that I have not mentioned the percentage of non-presented non-target words in this paragraph. I want to use this to make a point about schematic memory in the next paragraph.] Whether false memories such as those demonstrated by Roediger and McDermott (1995) result from the same processes as those arising from episodic memory has been questioned by Freyd and Gleaves (1996).

They argue that there is little similarity between a false memory of a word that did not appear on a list and a false memory of childhood sexual abuse. [Note that we have come full circle here. We set out what the debate was and offered two viewpoints. Having examined one in detail we have returned to the other.] Furthermore, while it is possible to create a false memory of being lost in a shopping mall, a false memory of having had a rectal enema is difficult to achieve (Pezdek et al., 1997). This may be because this type of event is too dissimilar from existing real memories. Support for this explanation is from the fact that in the Roediger and McDermott (1995) study, only 2 per cent of non-presented unrelated words were falsely recognised. Since these words were semantically unrelated, the processes involved in false memory may be based on semantic content. Thus a false memory for a non-presented word in the lab may occur because semantic information may be organised schematically.

Indeed, a false memory may be the result of schematic memory and inferential processes. In other words, a false memory may only occur when it is first related to an actual experience and second whether it is plausible or not. [The discussion has now moved on to how false memory could be explained using our knowledge of models of semantic memory. So, you could extend this discussion by examining several theories of semantic memory and asking whether each or any can account for these findings – see Chapter 5: Semantic Memory. The next paragraph is one example of how to do this]. Much of the data from false memory experiments can be interpreted as examples of schema-driven errors in memory rather than the creation of false memories in the sense of recalling an episode or event in life that did not happen. [Insert a brief description of what schema theory is here.] For example, in a study by French and Richards (1993), participants were asked to draw a clock with numbers represented by Roman numerals.

Some were asked to draw the clock after having been shown the clock for six minutes, and others copied from a drawing of the clock in front of them. They found that only participants in the copy condition drew the clock without any errors. That is, they drew the numeral representing four as IIII and not as IV. [The Roman numeral for four is nearly always written as IV, except on a clock face where it is drawn as IIII for aesthetic reasons.] This demonstrates that a schema in memory can under certain conditions overwrite what one sees. However, this would not be regarded as a false memory but results when information is schematically organised in memory. Hence experiments like Roediger and McDermott's may be better understood within the framework of schema theory. [All that is left now is a coherent conclusion, and a correctly constructed References section].

Many people claim to have recalled a so-called long-lost memory. It may be that it is an accurate memory of a real event. However, research has shown that false memories can be created whereby the individual believes the memory to be true but it is false. In addition, research shows that confidence and the reality of the memory do not correlate. Therefore, the recovered memory and the false memory may both be partially correct. The research on this topic has shown that human memory is flexible and malleable, and it is open to making errors. These errors are not just about how information can be forgotten but also about the construction of events that never happened.

Question 3: Hypothetical student report on a form of implicit learning

TITLE: Evaluative learning: affective stimuli and the occurrence of implicit learning [The title gives an indication of the topic, but should also mention the variables used].

ABSTRACT

Evidence suggests that evaluative learning (a procedure for changing the feelings towards stimuli) can occur independently of conscious awareness (when the participant is unaware of the change). In this experiment, liked and disliked stimuli were presented subliminally (too brief to be consciously detected) and immediately before neutral stimuli. Participants were told about the expected influence of the affective stimuli on the neutral stimuli in terms of how they would rate them. They were further instructed to resist the effect of the subliminal stimuli when rating the neutral stimuli. The findings provide further support for implicit evaluation learning. [Notes: The first sentence introduces the topic area and the main research question. The abstract then describes a complex procedure in a couple of well-structured sentences. It ends with an unambiguous conclusion. An indented paragraph is visually pleasing but not necessary. A sentence about the actual results is missing, such as ‘The main result was that participants were unable to resist evaluative learning when instructed to do so.' Check with your tutors about writing abstracts, as some colleges now expect you to use structured abstracts.]

INTRODUCTION

Understanding our emotional responses to everyday objects and situations enables us to appreciate the ways in which we acquire our attitudes (Zajonc, 1984; De Houwer et al., 1994). According to Osgood et al. (1957) we attach a value to all such experiences, thus forming our preferences. It has further been argued that our preferences may be made without our cognitive awareness (Zajonc, 1980;Murphy and Zajonc, 1993). Understanding this phenomenon is of great importance to our knowledge of human behaviour. [Notes: The introduction begins with the issue of preferences. There are no general and really obvious statements like ‘Memory is really important, without it we would not be able to remember anything'. Next, the issue of awareness is introduced immediately. This way the reader is told straightaway what the research concerns. A definition of evaluative learning follows. It is important to begin to define very important terms as soon as possible.]

Much evidence has been put forward to support the theory that we may develop our preferences through evaluative learning (Levey and Martin, 1975; De Houwer et al., 1994; De Houwer et al., 1997; Johnsrude et al., 1999; Fulcher and Hammerl, 2001). This behavioural paradigm proposes that when a neutral stimulus is paired with an affective stimulus, a shift in valence (change in liking or disliking) will occur towards the neutral stimulus. An everyday example might be disliking the name Peter because of previous bad experiences with someone of that name. Levey and Martin (1975) demonstrated this by presenting participants with a number of pictures and requesting they be sorted into three categories; liked, disliked and neutral. A neutral picture (conditioned stimulus) was paired with either a liked or a disliked picture (unconditioned stimulus). When the valence toward the neutral picture was measured, a neutral picture paired with a liked picture was also liked. In contrast, a neutral picture was disliked if it had previously been paired with a disliked picture, thus supporting the theory that we acquire such preferences through evaluative learning. [Next, the author introduces the main opposition to this theory.]

A growing body of evidence suggests that evaluative learning can occur independently of our conscious awareness (Lewicki et al., 1992; Hammerl and Grabitz, 1996; Fulcher and Cocks, 1997; Fulcher et al., 2001). For example, one might still dislike the name Peter even though one has forgotten about that person and the associated bad experiences. The modern approach to investigating this phenomenon has extended the initial work of Martin and Levey (1975). Recent evidence confirms the occurrence of evaluative learning when liked and disliked stimuli are presented subliminally (too briefly to be consciously detected) and, therefore, outside the realms of consciousness (Krosnick et al., 1992; De Houwer et al., 1997; Fulcher and Hammerl, 2001). In contrast, other researchers reject this proposal and suggest that the concept of learning is only related to awareness (Cohen, 1964; Holender, 1986). Further evidence in support of this argument has been presented by Klein and Wiener (1966) who confirmed that lowered awareness levels did not result in learning. Recent evidence (Field, 2000a; Field, 2000b) also disputes the concept that learning can occur without awareness and more specifically critiques the methods of data collection employed by those researchers who support this claim. [Next, we are told how proponents of the main theory have responded to such criticisms. Note that this response is the first hint of the actual method to be used in the current study.]

In response to such criticism, Fulcher and Hammerl (2001) varied the length of time affective stimuli were presented. In this instance, pictures of faces expressing emotions of anger or happiness were presented subliminally (for 12.5 ms) or overtly (for 125 ms). Immediately afterwards, Chinese ideographs (Chinese symbols) were presented. Participants were asked to rate whether each ideograph was ‘liked' or ‘disliked'. The results confirmed that subliminal presentation of stimuli had resulted in implicit evaluative learning because those ideographs presented with a happy face were rated as ‘liked' and ideographs were ‘disliked' if they had previously been paired with an angry face. [Having described the published study that is closest to the current experiment, and in some detail, the next paragraph (a clear statement of the rationale for the current experiment) is easy to write.]

The purpose of this experiment was to replicate the method of Fulcher and Hammerl's (2002) study but by revealing the nature of evaluative learning to participants in advance of the experiment and by including the instruction to resist the effect when subsequently rating the neutral pictures. It was predicted that participants would not be able to resist evaluative learning as indicated by their ratings of the neutral stimuli. [Note the way the hypotheses are presented. They are not presented as ‘Hypothesis 1: There will be ...', but rather in the form ‘It was predicted that ...'. The former style is acceptable, but if you read academic journals you would know that the latter style is most commonly used (but check the advice from your department as they may insist on a particular style for stating the hypothesis or null-hypothesis). Other important points to note about this introduction:

•  it does not waffle on about studies that are tangential (not absolutely essential) to the current study;

•  it is brief and written without ambiguity, in good English, and to the point;

•  there is evidence that the student has read the recommended reading;

•  students who do not read the recommended reading for an assignment can't possibly hope to write well on the subject – and they certainly do not fool the tutor when they try to change the debate or talk about issues not directly related to the subject.]

METHOD

Design:

The experiment employed a between subjects, repeated measures design. The independent variable was whether the affective stimulus was of a smiling or an angry face. The dependent variable was the ratings of the ideographs. [Note: The design of the study is clearly stated, as are the independent and dependent variables. Always begin the Method with a word about the design. State what type of design it is, what the IV is, and what the DV is. State these accurately].

Participants:

Sixty-eight participants took part in this experiment, 80 per cent of whom were female. The overall calculated mean age was 24.28. Equal numbers of participants were assigned to each condition (n=34). [Note: Don't provide unnecessary details about the participants. You could add that they were psychology undergraduates enrolled on a cognitive psychology module, but not much more is needed.]

Apparatus:

The experiment took place on the computer. All experimental instructions were presented to the participants on the computer screen and the experimenter confirmed that these had been fully understood. Participants were requested to select the ‘A' key on the computer keypad if they considered an angry face had been presented and the ‘5' key on the numeric keypad if they considered a smiling face had appeared. [Note that the author does not provide a list or a bullet pointed list. The apparatus or materials need to be discussed using complete, grammatically correct sentences. The author could have mentioned that digitised ideographs, and digitised images of smiling and angry faces were used.]

Procedure:

Individuals were requested to pose for photographs exhibiting either a smiling or an angry face. At random, the photographs were paired with Chinese ideographs. The photographs appeared on the computer screen for 12.5 ms. Immediately afterwards, an ideograph was presented and remained on the computer screen for two seconds. Participants were told that subliminal faces would be presented and the possible effects of this. [Note: It is important to mention randomised presentations and counterbalancing where they occur. This is hinted at above but not discussed in enough detail.] The experiment comprised three phases. Phases one and two requested participants to simply rate whether they had seen an ‘angry' or a ‘smiling' face. Phase three asked participants to resist this effect, while they rated them on the scale 710 to +10. [Note: The procedure is quite well written: it is clear, logically structured and to the point. The author could state that 710 is the highest dislike rating and +10 is the highest like rating, and that the values increase from 710 in units of 1. The instructions are summarised rather than printed in full (with severe word limits, you have to summarise them, but without such limits you can print them in full if you think it would help the reader understand the procedure, which it often can. Again, seek advice from your tutors on this issue as they may disagree.]

RESULTS

The mean ratings for ideographs paired with smiling faces and ideographs paired with angry faces were calculated, and can be viewed in Table 1. [Note: When you create a table always refer to it in the text, as here. Don't let it ‘speak for itself'.]

The data was analysed using a related t-test. The results confirmed that there was a significant difference between mean ratings for ideographs paired with smiling faces and ideographs paired with angry faces (t (66) = 5.255, p <0.01, computed p = 0.0034). The effect size was calculated as 0.87, which means that this procedure produced a strong effect. [Note: The results are reported within a meaningful sentence. Don't have it printed on a line by itself nor should you say something like: ‘The results were significant, t(66)= 5.255, P < 0.05' without saying where the difference lies. Remember you are not reporting the results of a football match. Tottenham 5 Manchester Utd 0 speaks for itself but t(66) = 5.255, P < 0.05 does not. It is the difference between the means that is statistically significant. The effect size tells us that there is a strong effect and that statistical significance would have still been reached with much fewer participants.]

Table 1 Mean ratings of ideographs paired with smiling or angry faces [Note the table should benumbered and have a title. The title should mean something to a reader who is just scanning the report. Two decimal places would suffice for data like this. The student could have also stated that the standard deviations for each condition are in brackets.]

DISCUSSION

The results of this study support the prediction that there would be a significant difference between the mean ratings of ideographs paired with smiling faces and that of ideographs paired with angry faces. Thus first confirming that learning can occur without our conscious awareness, since the faces were presented outside of conscious awareness. [Note: The first thing you do in the discussion is to relate the results to the hypothesis.] Second, these results occurred in participants aware of the evaluative learning effect and who were trying to resist it. More specifically, the results conclude that evaluative learning is independent of contingency awareness and is impervious to conscious control. [Note: The author takes the first step in identifying the implications of the results. Note that the term ‘the results conclude' is inaccurate, and should say something like ‘from Mean ratings of ideographs paired with smiling faces Mean ratings of ideographs paired with angry faces 6.2794 (1.94) 4.4559 (2.15) these results it is possible to conclude'. Next, the author discusses the results with respect to the main theory.]

The conclusions of this report replicate the recent findings of Fulcher and Hammerl (2002), thus providing further evidence for their argument. Arguably such conclusions confirm the occurrence of implicit evaluative learning. In other studies, before the pictures are presented as pairs, the participants rate how much they like or dislike each neutral picture and again after the pairings. It is found that neutral pictures paired with pleasant pictures show an increase in their rating, while neutral pictures paired with unpleasant pictures show a decrease in their rating (Levey and Martin, 1975). In this study, ratings were taken at the end only, and a comparison of the ratings of ideographs paired with smiling faces against the ratings of ideographs paired with angry faces taken. Although such differences may have occurred through evaluative learning there may be an unlikely, but still possible, alternative explanation, such as poor counterbalancing. [Note: Next, the author discusses the wider implications of the results and adds some caution.]

In a wider context, the finding reported here is an important contribution to our understanding of the ways in which we acquire our preferences. It has been suggested, however, that generalising conclusions, such as those presented in this study, outside the laboratory must be done with caution (Rozin, et al., 1998). It is, therefore, proposed that future research could focus on the study of preferences in the real world, thus increasing the ecological validity of evaluative learning. [Note: There is a discussion of a possible future study but more details are required to take this suggestion seriously. The author could provide an outline of such an experiment. There aren't many criticisms of the method. Always think about the materials used and why they may have been chosen. Are they representative of materials found in everyday life? Is the procedure tightly controlled? A positive feature is that the author does not make trivial criticisms or unsupported statements, such as ‘all Ps were psychology students, therefore we cannot generalise to the population' – why not? Or ‘there were too few Ps to draw any firm conclusions' – how many should we have then? How many is too few? To make comments like these we need some supportive evidence. For example, if you are to consider the issue of the number of participants, you need to discuss the issues of ‘statistical power' and ‘effect sizes'. When you have calculated these you can then say whether you needed more Ps or not.

Overall, the student has adhered to the conventions of report writing, but has done so in a way that provides clear evidence that she is familiar with the format of journal articles. This correct formatting, her level of understanding of the central issues and the quality of her writing style is clear evidence that she has read the recommended reading and shows a good level of understanding. She also demonstrates the ability to critically evaluate arguments and understands the limits of empirical research. Note that the references should follow but I have not included them here as they are included in the general references section on the next page.]

Next >>

This book was first published in 2003 by Crucial, a division of Learning Matters Ltd [ISBN 1 903337 13 5] © 2003 Eamon Fulcher; © 2009 GEFT Consultance Services (geft.co.uk).

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without prior permission in writing from Geft Consultancy Services, who may be contacted via www.geft.co.uk.