Murdock conducted a procedure very similar to Glanzer and Cunitz although he only used one condition; none of his participants were given a distracter task.
He found that nouns, which were presented towards the beginning and the end of the list, were recalled greatest and recall of words in the middle of the list tended to be less. From his findings he was able to draw the conclusion that when individuals recall information from a list, such as the list of nouns he used, they recall the information from both STM and LTM.
Pieters and Bijmolt (1997) investigated consumer memory for television advertising. This study holds great relevance to the study I will be carrying out as it concerns advertising. In this field experiment the researchers collected data from the Netherlands after a certain block of adverts was broadcasted between a certain television programme.
The viewers were interviewed 30 minutes after they viewed the adverts and were asked to free recall what they had seen. The findings showed evidence to support primacy effect, as the adverts towards the beginning of the block were recalled most successfully. There was some evidence to support recency effect but this was only due to the fact that recall for the advert shown at the end of the block was comparatively better than that of the adverts shown immediately before.
In a more controlled laboratory experiment, Scott (2005) studied into the effect serial position has on recall of television advertisements. This study most closely related to the experiment I will be carrying out myself.
Participants in Scott’s study were asked to view 16 adverts of which the presentation order was randomised. The findings of this study showed that recall of adverts within the positions 1-3 were recalled most consistently (primacy effect), similarly with positions 13-16 (recency effect). Recall of adverts in the middle of the series was less consistent (positions 7-9). However, he, alike with Glanzer and Cunitz (1966) found that if recall was delayed, the recency effect disappeared. Following the results of Scott’s study, it was suggested that the most effective time to broadcast a television advertisement is at the beginning of the block, as his results implied that it is at this time when the advert will have greatest impact on the viewers’ memory.
Formulation of Aims: To investigate the effect serial position has upon recall of magazine advertisements. To discover if primacy and recency both play a part in recall of adverts i.e. are STM and LTM utilised when recalling information of this nature.
I have adapted Scott’s (2005) study by allowing participants longer to view the adverts. I have decided to do this to ensure participants have sufficient time to view adverts and to eliminate the chance that participants forget certain adverts as a result of lack of time to view the adverts rather than a result of primacy or recency effect.
I expect to find that adverts at the beginning (positions 1-4) and the end (positions 13-16) of the block of adverts will be recalled most, in comparison to the adverts in the middle (positions 5-12) of the block.
Directional Hypothesis: Products from adverts at the beginning (positions 1-4) and end of the block (positions 13-16) will be correctly recalled significantly more often than adverts in the middle (positions 5-12) of the block. I have identified this directional hypothesis because other studies have shown that information in the primacy and recency positions increases the likelihood of its recall.
Null Hypothesis: The serial position of adverts in a block of 16 will have no effect on recall of the products from these adverts.
Method
Experimental design: This experiment was a laboratory experiment as it eliminated the risk of confounding and extraneous variables affecting the results. For example, controls were put in place to reduce the chance of the participants’ concentration being interrupted. Other advantages of this method are that the experiment was more reliable as the controls ensured it was replicable and cause and effect can usually be claimed by using this method of testing.
Although, the disadvantages are that laboratory experiments can be criticised for not being valid as the controls make it unnatural and it is usually only the independent and dependent variables that are examined; other variables are rarely taken into account.
This experiment utilised repeated measures design as it was the most valid design to use because individual groups design was not possible due to the nature of the experiment (all participants saw a beginning, middle and end of a series of adverts). Another advantage of this experimental design is that fewer participants are needed which would make the group more manageable; it also reduced the risk of participant variables. For example, some participants may have better memories than others.
However, disadvantages of this experimental design include the introduction of order effects. For example, participants may be better or worse at the task due to practice or fatigue. Counterbalancing, however, can prevent order effects from becoming a big problem. Counterbalancing was put into effect by differing the order of presentation of the adverts and ensuring the adverts were collected from a variety of sources (participants were asked to recall the products in the order they believed they were presented). This reduced the risk of some adverts being recalled more or less due to their distinctiveness.
Independent and dependent variables: In this experiment the independent variable was the serial position of an advert, either at the beginning and end or middle of a block of 16. The beginning and end of the block are defined as positions 1-4 and 13-16. The middle of the block is defined as positions 5-12.
The dependent variable was how many products were recalled successfully by participants, out of the potential 16 adverts.
Possible confounding variables:
- The adverts chosen may have had certain personal significance to some.
Solution: Select adverts from a variety of sources and which were similar in content and layout. This would be done by establishing some standard method of comparing different adverts. For example, categorising each advert into certain groups, depending on the size of the adverts, the topic the advert addressed and the number of words printed on the advert.
- Distractions may have affected concentration; causing reduced recall of the advert shown at the time the distraction occurred.
Solution: The experiment was carried out as a laboratory experiment so the controls ensured there were no interruptions (a No Entry sign was placed on the classroom door).
- Some participants may not have been able to see the adverts clearly.
Solution: the participants were asked to move to a position where they had a good view of the adverts being shown.
Ethical considerations:
-
Participants were briefed thoroughly before beginning the experiment, once the participants understood what the experiment would entail; it was possible to obtain their informed consent. Deception was not used.
-
Participants were allowed the right to withdraw from the experiment at any time during the procedure.
-
Participants were not subjected to any humiliation by forcing them to recount what they have recalled aloud or in front of their classmates.
-
Participants wrote down what they recalled on a slip of paper, anonymously, thus there was no need for individuals’ observations to be revealed to others. The results remained confidential.
-
Participants below the age of 16 were not allowed to take part in this experiment unless they had parental consent.
-
All the participants were debriefed afterwards and it was ensured that they all felt comfortable they had not been subjected to any humiliating or uncomfortable circumstances during the experiment.
Participants:
- 20 participants in the experiment.
-
The target population was teenagers and my target sample was students in the sixth form as they are 16+. I had to use sixth form students due to ethical limitations.
-
The sampling method used was opportunity sample because a class teacher had given permission for their AS level class to be used and therefore participants who were in the classroom at the time I had planned to carry out the experiment took part, with their consent.
Equipment:
- A set of standardised instructions (see appendix 1)
- A set of 16 adverts taken from a variety of magazines (see appendix 4)
- Paper for participants to write down what they recalled to keep the results anonymous.
- Pens for the participants to write with.
- A debrief (see appendix 3)
Procedure:
- The participants, who were already seated in the classroom, received the standardised instructions as a whole group.
- The participants were asked if they were comfortable with the procedure and were told that they could leave the experiment any time if they wanted.
- The class of 20 was split evenly into two groups randomly by selecting a slip of paper from a bag which said whether they were to be part of Group A or B.
- The participants of Group A then viewed the 16 adverts one at a time for 15 seconds each.
- They were then asked to recall the product that was advertised in each advert in the correct order.
- Their answers were recorded on individual sheets of paper.
- The experimenter collected in the sheets.
- The participants of Group B viewed the same 16 adverts for 15 seconds each but the order of presentation was different to Group A’s. The order was changed in such ways as to ensure that adverts, which were shown at the beginning and end in the first viewing, were shown close to the middle for the second viewing.
- Participants recorded their answers again on sheets of paper in the order they remembered seeing them.
- The experimenter collected the sheets in.
- The participants were debriefed and were given an opportunity to ask any questions.
- Questions were answered and participants were dismissed.
- Participants’ answers remained confidential.
- The results were taken away for analysis.
Results
A table to show the measures of central tendency in an experiment where lists of adverts were shown to 20 participants who were then asked to free recall these adverts
Inferential statistics:
The Wilcoxon T test was selected for this type of data because the experimental design was repeated measures and the data, which was collected, was ordinal as the scores of recalled adverts were ranked. Also, the hypothesis stated there would be a difference between recall of adverts in the primacy and recency and adverts in the middle.
- The level of significance that must be reached for the hypothesis to be accepted is 5%.
- The calculated T score (4.5) must be either equal to or less than the critical value (60).
- 4.5 is a much lower value than 60, so therefore it is appropriate to accept the experimental hypothesis that adverts at the beginning and end of the block will be correctly recalled significantly more than adverts in the middle of the block, because the calculated T score is significantly lower than the critical value. I can, therefore, accept the hypothesis with reasonable confidence.
Discussion
Explanation of results: The directional hypothesis in this experiment stated that products from adverts at the beginning and end of the block would be correctly recalled significantly more than adverts in the middle of the block.
The graph in the results section suggests that this hypothesis is correct as condition A (primacy/ recency: positions 1-4 and 13-16) clearly shows higher central tendencies (mode - 7, median - 7 and mean – 6.7) than condition B (middle: positions 5-12) which displayed central tendencies of mode – 5, median – 5 and mean – 4.6.
After calculating the Wilcoxon T score, it is possible to see that the T score (4.5) is a much lower value than the critical value (60). This means, therefore, that the results of the experiment display that there was more than 5% difference between condition A and B. This 5% was the significance boundary that was selected as the experiment focussed on an average test of memory which, as a result, did not require a significance boundary that was too lenient or stringent. Therefore, I have accepted the Experimental hypothesis and rejected the Null hypothesis. Due to the results showing a significant difference between the two conditions it is acceptable to reject the Null-hypothesis, which was that there would be no difference at all between the amounts of products recalled from both conditions.
Relation to previous studies: The results of this experiment agree with the findings of all the experiments which were discussed in the introduction (Glanzer and Cunitz, Murdock, Pieters and Bijmolt and Scott).
Although Pieters and Bijmolt’s study presented evidence to support the primacy/ recency effect, the procedure allowed participants a delay before recall, which makes the results of this experiment less relevant to the study described in previous sections. Also, their study was carried out as a field experiment, which means that although the ecological validity would be high, the reliability and validity would be low as it was difficult to control the environment and variables.
The results of Scott’s (2005) laboratory experiment relates most closely to the findings of this experiment as he found that recall of the adverts in the primacy and recency positions were recalled most consistently, and adverts in the middle of the block were recalled less consistently. The results of this experiment support the hypothesis that primacy and recency effect does affect recall of information in those positions in a list.
Limitations and modifications: One potential limitation with this study was that some participants used word association techniques in order to remember more of the adverts. This may have affected the results and made them less valid. If the study was to be carried out again, the standardised instructions would have been modified so the participants were told not to use any techniques that might help or hinder their recall of the adverts.
Another possible limitation was that the experiment was carried out at the end of a lesson. This could have meant that some participants were tired or bored of their surroundings, resulting in them feeling less motivated to concentrate and recall the adverts to the best of their abilities. A modification would be to carry out the experiment at the beginning of the day when participants may feel less fatigued.
Also, some adverts may have been more distinctive to some participants than others.
To improve this, participants who are all exposed to different media genres could be used to produce generalisable results for a wide variety of people.
Implications of the study and suggestions for future research: This study into serial position is significantly ecologically valid because it may relate to commercial advertising. This study suggests that adverts at the beginning and end of a block are better remembered than studies in the middle of a block.
This study also may hold some relevance for education. The study could support the argument that material that is taught in lessons at the beginning and end of the school day may be remembered significantly more consistently than material taught in the middle of the day.
Further research could be carried out to investigate the theory that the order of lessons in a school timetable may affect children’s ability to learn the subject.
A simple test could be carried out at the end of each school term in each of the subjects and results will be analysed and referred to the position of the lesson within the school day.
Appendix 1
Standardised instructions:
Thank you for consenting to participate in my experiment.
I’d like to make it clear that you all have the right to withdraw from this experiment at any time if you feel uncomfortable.
If you feel threatened in any way while participating, the experiment will stop immediately.
I am going to ask you, in a moment, to pick a slip of paper out of a bag. On these slips of paper you will see either a letter A or a letter B. These letters refer to the group you will be in.
After selecting your letter I will split the class accordingly and show Group A a block of 16 magazine adverts. These participants in Group A will then be asked to write down the adverts they can remember in the correct order, or as correct as possible.
I will then collect in Group A’s answers and repeat the same process with Group B.
All your answers will remain confidential.
I will answer any questions you may have at the end of the experiment.
Appendix 2: Raw Data
Order of Presentation one
Group A
Primacy and recency = positions 1-4 and 13-16
Middle = 5-12.
Order of presentation two
Group B
Primacy and recency = 1-4 and 13-16
Middle = 5-12.
Wilcoxon T score = 4.5 (calculated T score) because it is the lowest of all the ranked scores.
Numbers in red indicate the rank number.
Numbers in blue indicate the difference number.
Appendix 3
Debrief:
First of all, I would like to thank you for participating in this experiment.
As you may have already guessed, the aim was to investigate the effect the position of an advert has upon your memory.
The second part of the aim was to find out if adverts that were at the beginning and end of the block were better remembered than those in the middle. You might have realised that the adverts shown to you that were in the middle of the block of 16 were the most difficult to remember.
Both group A and group B were shown the same adverts but in different orders.
The information I have taken from this experiment will not be used in any other way, besides helping me interpret the findings of this investigation.
Appendix 4
Adverts:
References
Glanzer, M and Cunitz, A.R (1966): ‘Two mechanisms in free recall’: Journal of Verbal Learning and Verbal Behaviour, 6,928-35.
Murdock, B.B. (1962): ‘The serial position effect in free recall: Journal of Experimental Psychology, 64, 482-88.
Pieter, R.G. M and Bijmolt, T.H.A (1997): ‘Consumer memory for television advertising: A field study of duration, serial position and competition effects’: Journal of Consumer Research, 23,362-72
Scott, T.W. (2005): ‘Serial position effects in recall of television advertisements’: Journal of General Psychology, 132.2, 151-63
