Scoring methods and it's directional effect on dementia or depression diagnoses

Comments · 44 Views

Research has shown a link between serial position effects (recency and primacy) and dementia and depression. Mild cognitive impairment/early dementia is linked with a pronounced recency effect and an impaired primacy effect, while depression is linked with an impaired recency effect and a


Differences in the types of memory deficits can be seen for several different reasons, for example, different medical conditions. People with certain conditions can show drastically different memory deficits, for example, people with depression or a neurological disorder such as Alzheimer’s disease. These are two conditions that are being compared in this study and these diagnoses can produce vastly different memory impairments. One way in which the memory deficits differ is the timing in which the memory decays. This timing of memory decay is called the serial position effect. The serial position effect is where words positioned at the beginning of a list (primacy) and at the end of a list (recency) tend to be better recalled than words positioned towards the middle of a list. When these scores are plotted on a graph this shows a U shape curve. In patients with cognitive impairments from a neurological condition, such as Alzheimer’s disease and depression, this shape is seen. It is important to control for both of these conditions when studying the serial position effect as it may help diagnostic efficiency when attempting to differentiate Alzheimer’s disease and depression.

Memory and Depression

Depression is a psychological condition that affects the cognitive aspects of a person’s life. It has been explained simply as the lowering of mood or spirits (Simpson and Weiner, 1989) but the effects of depression can be far more wide reaching than this, for example impaired attention and memory. Depression is becoming an increasingly common issue among people, with the World Health Organisation (WHO, 2017) believing there is an increase of more than 18% of people suffering from depression between the years 2005 and 2015. A study by Lim et al. (2018) estimates that the lifetime prevalence rates of depression among people in a sample of 30 countries is approximately 10%. There are numerous conditions that affect memory in those diagnosed with a degenerative neurological impairment. Those diagnosed with such a condition typically experience progressive memory deterioration and deficits in their executive cognitive abilities. Certain conditions deteriorate different parts of the brain which are associated with the different types of memory. It has been long theorized that depression affects people’s cognitive abilities negatively (Potter et al, 2004). Cognitive impairment is a (ICD 11) diagnostic criteria for single and recurrent depressive episodes without psychotic symptoms. Many people will at some time will be mildly depressed without psychotic symptoms in a single isolated episode, whereas the people in this study are diagnosed with recurrent depressive episodes and that may experience psychotic symptoms. People who are depressed have been shown to have memory deficits when compared to healthy controls (Rock et al., 2014). Studies have shown that depressive episodes include deficits in the executive functions of short-term, long-term and working memory (Savard et al. 1980; Mormont, 1984; Rubinow et al., 1984; Castaneda, et al., 2008).These studies measure broadly if there are memory deficits using total scores or subscale scores, comparing the results to healthy controls and less severely depressed individuals, our study proposes to compare these results against cognitively impaired participants to examine the similarities of depression and severe neurological disorders. The use of total numbers of words recalled may not be the most accurate means to assess memory scores especially in older adults, even though it is recommended in the manuals accompanying memory assessments. One principle reason that the overall memory score approach may not be the most accurate is that it gives equal score weighting to items that may be assessing verbal attention and not memory. The nature of the condition that is in focus for this assessment may highlight the poor fit of this measure. For example, people diagnosed with Alzheimer’s disease will develop memory problems and often this problem comes to prominence as an episodic memory problem, while verbal attention may not be notably impacted by the disease. Therefore, we would expect that this would be recognized as a reduction in the primacy effect, but the recency effect preserved, in relation to the immediate recall performance. This may not be the case of items are equally weighted as it may inaccurately score items that are not significantly impacted by the disease.  Individuals who are depressed show difficulties with tasks that require sustained efforts, such as list learning or free recall tasks. Research has suggested that there is an inability to acquire new information in majorly depressed individual without any deficit in one’s ability to retain already acquired information (Cronholm Ottosson, 1961). This inability to acquire new or recent information is theorized to be a result of the intrusive thoughts that depressed patients experience and result in the decreased ability to acquire new information (Sternberg Jarvik, 1976; Stromgren, 1977). Research by Richard et al. (2013) cross-sectionally associated mild cognitive impairment with depression. Majorly depressed individuals are also shown to be the most severely cognitively impaired.

The serial position effect is a psychological phenomenon that is seen in cognitive psychology and memory, and it was first noted by physicist Francis Nipher in 1878. This effect is when the words recalled in a list learning task are more easily recalled in the beginning and end of the list (Troyer, 2011). An example of this may be out of a 9-word list, participants would usually recall approximately 70% of the first three words and 60% of the last three words, while only recalling 40% of the middle three words (Troyer, 2011). The primacy effect is reduced in individuals with memory disorders such as Alzheimer’s disease (Troyan, 2011). With reference to neuroimaging research, Medial temporal brain regions are more active for the retrieval of initial list items (primacy effect) and this illustrates the long-term memory involvement in the primacy effect (Troyan, 2011), this may explain why the primacy effect is reduced in individuals with Alzheimer’s disease. As the diagnosis of dementia is still mainly based upon its clinical criteria (Staffaroni et al., 2017) and Alzheimer’s disease is a form of dementia, these conditions contain overlapping symptoms and thus would make an accurate diagnosis of either more difficult. The serial position curve for a patient with Alzheimer’s disease will form a particular shape. Supported by the research above it will show a reduction in the primacy effect and an intact or exaggerated recency effect (Fodi et al., 2003). In individuals with normal cognitive function, words at the beginning and end of the word list are more frequently and more easily recalled than words in the middle (Deese Kaufman, 1957). These patterns of word recall are known as primacy and recency effects. Traditionally, the serial position effect is graphed as a U- shaped curve and this consists of two components that explain this phenomenon, the primacy and recency effect. The primacy effect shows an advantage in recalling information that was presented at the beginning of a list, this advantage in processing is given to items first in a list and is then selectively transfers these items into long-term memory (LTM). The primacy effect relies on the information being transferred to LTM because of the time elapsed between reading the stimulus and recall is too long to rely on short-term memory. Long-term memory has been long traced to the hippocampus region in the brain, the part of the brain responsible for long-term potentiation, damage to this area of the brain would result in long-term memory deficits similar to that seen in Alzheimer’s patients. This is because damage to this part of the brain containing the hippocampus would negatively impact its structural integrity and would reduce the efficiency of the process involved in long-term memory storage. However, concurrent with research memory is not a singular brain region but a collection of cognitive processes that are produced in numerous major neural systems (Dickerson Atri, 2014). The recency effect is the advantage given to processing information at the end of a list, this system of retention relies on the short-term memory store (STM). Short-term memory is found in the frontal lobes and would also result in memory deficits but rather in short term and working memory. Research by Burkart, Heun Benkert (1998) shows that in elderly patients with diagnosed dementia, there is an immediate recall recency effect observed in these individuals. While a definite cause for the recency effect has not been disclosed, there is evidence to suggest that patients with Alzheimer’s disease exhibit preservation of the recency effect because their short-term memories are less affected by the disease. Many of these researchers have focused on the consequence of a diagnosis of Dementia on cognitive ability, some have taken the biological consequence of these diagnoses into account. Research has provided evidence for these models by examining people with damage to specific parts of the brain and studying their memory capabilities, finding that people with damage to parietal and temporal lobes exhibit impaired short-term phonological capabilities while their long-term memories remain relatively untouched (Shallice Warrington, 1970; Vallar Papagno, 2002). By contrast, patients with damage to the medial temporal lobe are seen to have reduced long-term memory capabilities while their short-term memories remain intact (Baddeley Warrington, 1970; Scoville Milner, 1957). This research pinpoints the biological effect of these degenerative diseases on the brain and increases our understanding of these brain regions and their role in memory. There does not even need to be damage to these parts of the brain to result in significant cognitive impairment, a study by Beblo et al. (1999) studied the neuropsychological correlates of major depression. In it, patients with depression were investigated, the researchers concluded that consistent with the statement of Veiel (1997, P. 587) stating that the deficits seen in depressed individuals are consistent with impairments seen in individuals with frontal lobe deficits. These findings cannot be generalized however, due to the small sample size of the study.

Damage to one specific brain region can interrupt cognitive function and produce memory deficits. For example, hippocampal damage due to damaged axons and dendrites as a result of the onset of Alzheimer’s disease, this hippocampal damage leads to severely impaired long-term memory issues (Scheibel, 1983). A famous case of hippocampal damage is the case study of Henry Molaison (H. M.). H. M. had his hippocampus removed and suffered extreme anterograde and retrograde amnesia, while his working memory remained intact, highlighting the importance of the hippocampus in episodic memory. While this severe brain injury is rare, it mimics the memory deficits seen in Alzheimer’s disease even when the physiological damage is not identical, this is evidence for the hippocampus being essential for long term memory. Hippocampal interference and degeneration are an important aspect of Alzheimer’s disease and research has shown that this leads to severely deteriorated episodic memory (Salmon, 2000). A study by Cunha et al. (2012) examined the serial position effect in individuals with Alzheimer’s disease when compared to people with Mild Cognitive Impairment (MCI). This study examined the primacy and recency effects and compared the results of the two groups. The results showed that patients with Alzheimer’s disease have a severely impaired primacy effect compared to a recency effect. This finding supports the belief that a diminished primacy effect is a defining feature of Alzheimer’s disease (Cuhn, 2012). It is not only neurodegenerative disorders that can affect cognitive functioning and more specifically memory. Depressive episodes and depressive disorders have shown to impact memory. Research has shown that people suffering from depression have significantly reduced cognitive abilities and their memory is impaired when compared to healthy controls (Baune et al., 2010). More specifically, the memory impairments seen in majorly depressed individuals are primarily working memory and short-term memory deficits (Halvorsen et al., 2012). Major Depressive disorder has also been shown to impair the functioning of the frontal lobes. Furthermore, showing structural and functional problems with the Prefrontal Cortex (PFC) (Snyder, 2013). These structural and functional impairments are responsible for the impairment of executive functioning which includes short-term memory and working memory. Studies measuring frontal lobe activation and working memory tasks showed that depressed individuals have hyperactive frontal lobes (Harvey et al., 2005) this resulted in poorer performance in working memory tasks as participants with major depressive disorder were slower and less accurate (Walter, Wolf, Spitzer Vasic, 2007). This paper is attempting to combine these two separate areas of research to examine a relationship between depression, cognitive impairment and what effect these have on the serial position effect. This study will take patients with diagnoses of depression and of neurological degenerative conditions such as Alzheimer’s disease for example. We will then examine the interaction of these conditions with recall and recognition metric. Based on previous research, our study should find that participants with a depression diagnosis should show a primacy effect as research has shown that working memory and short-term in depressed individuals is impaired while long-term memory is intact. We also expect to find that in participants with a cognitive impairment diagnosis resulting from a neurological degenerative disease should produce a recency effect as research has overwhelmingly shown that long term memory is these conditions is severely impaired. Our research question is what effect does depression and cognitive impairment have on the serial position effect?




The aims of the study were to;          

Examine the data to discern whether the serial position effect has predictive potential when trying to identify if participants would eventually be diagnosed with dementia or depression.

  1. Research a directional relationship between the primacy effect and depression, seeing if depression presents with an impaired recency effect
  2. Observe the relationship between the recency effect and dementia, determining if dementia presents with an intact recency effect and a severely impaired primacy effect

The objectives were to:

  1. Recruit a sample of older adult patients who were being assessed for presence of early dementia vs depression
  2. Gain data on the serial position of items recalled from word list trials and other assessment parameters from these patients
  3. Examine the data in order to discern whether the serial position of these recalled words have value in identifying whether participants would eventually be diagnosed with early dementia or depression.


The principal hypothesis to be tested is that patients that were diagnosed with dementia will show a predominant recency effect in single word recall while patients with depression will demonstrate a predominantly primacy effect in single word recall.

Minor hypotheses - and state 3 or 4 minor hypotheses

  1. There will be a lessening of the serial position effect over the course of the 4 trials
  2. There will be an association between less formal education and a greater serial position effect in dementia but not in depression
  3. Differential scores between word list recall and story recall will not be significant in dementia but will be significant in depression.





A total of thirty participants took part in the study. The University Hospital Ethics committee approved ethics for this study.

            Participants who met the inclusion criteria (n=33) and participants who agreed to participate were referred to the regional HSE memory clinic in Limerick. These subjects were patients of this service and met the criteria for inclusion of this study.  Participants were aged between 65 and 84 (M =70.48, SD= 5.46), of these participants 13 (39.39%) were male and 20 (60.60%) were female. Participants education ranged from primary school education to University level education, in between these were group/inter certificate examination and the leaving certificate examination. Additionally, all of these participants were given neurological examinations and a battery of neuropsychological tests. It is worth noting that a significant portion of the participants left school following completion of their inter certificate, very few participants continued on to third-level education. These participants have been diagnosed with either depression or a mild cognitive impairment such as Alzheimer’s disease, dementia or some other form of neurological disorder. Cognitive examinations were administered by a licenced clinical practitioner to confirm diagnoses adhering to the ICD 11 diagnostic criteria. Participants have been referred to a memory clinic in the University Hospital, from where they are referred to multidisciplinary teams within this clinic which is made up of a range of clinical professionals. Each clinician assesses the patient and gives their professional input on a potential diagnosis, this information is then collated, and a diagnosis is made based on the evaluations of these clinical professionals. These diagnoses are given a diagnosis in this case of dementia or depression, so patients receive a diagnosis after the RBANS assessments have been completed, depressed or cognitively impaired, based on their diagnosis they will be included in a group. The patients were administered the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) test. The subtests of interest for this study are List learning, story memory, semantic fluency list recall, list recognition, story recall and figure recall (Randolph et al., 1998). These subtests were chosen because this study is examining memory function under different conditions and different presentations, these tests hence examine working, short-term and long-term memory. All participants received the same battery of neuropsychological testing, both groups of depression and cognitively impaired. The tests were administered on separate days over a prolonged period of time. Comparisons are made between the depressed and cognitively impaired participants. Subjects were diagnosed at different stages of cognitive impairment, for example, some cases of Alzheimer’s disease were more advanced than others.



All participants were given psychological examinations by a certified clinical psychologist belonging to the Psychological Society of Ireland (PSI). These participants were patients of the memory clinic service and were referred for clinical neuropsychological assessment as part of the memory clinic assessment protocol... Following assessments form the clinical professionals who were part of the multidisciplinary team patients subsequently received diagnoses:  either a Dementia or it a diagnosis of depression. All of these patients underwent the Repeatable Battery for the assessment of Neuropsychological Status (RBANS) (Randolph, 1998) test, which examines the cognitive function of an individual to attempt to detect potential cognitive impairment. All patients have completed full neuropsychological evaluations and consent was retrieved from all participating individuals. Ethical approval has been approved from the ethics board of the University hospital limerick research ethics committee. For both groups in this study, the RBANS test is the only form of psychological and cognitive assessment used.





The RBANS (Randolph, 1998) is an individually administered battery of cognitive tests measuring executive function in the areas of immediate and delayed memory, attention, language and visuospatial ability. The RBANS is made up of 12 subsets which converge to 5 index scores and a total score. These subsets were administered as intended by the manual for use and the scores were added up consistent with the intended scoring procedure. The RBANS assessment measures 5 areas, these consist of; immediate memory, visuospatial/constructional performance, language, attention and delayed memory. Each area has a number of subtests that combine to make a score for each area. Immediate memory consisted of list learning and story memory as subtests that measure immediate memory. List learning is a task where participants are required to freely recall items on a word list, there are four trials of free recall. Following this, participants move on o story memory, this is similar to list learning here participants are required to recall a story as accurately as possible after it is read to them. 

Moving on from immediate memory is visuospatial/constructional performance and this comprises of figure copy and line orientation. Figure copy is where participants are to replicate a drawing shown to them which is left on display for them. Line Orientation is where participants are presented with a line drawn at a specific angle and participants then must identify which line matches the presented stimulus.

Language is the next of the 5 assessment areas and this is assessed using picture naming. This is where participants are asked to name each picture presented to them, if a picture is misperceived then they are given a semantic clue to help.




The scores from the memory (This is where the memory subtests are listed) are scored based on their serial position, with this procedure in mind the retention weighted recall and recency ratio were employed to score the memory results. The retention weighted recall measure weights each recalled item in adherence with its serial presentation order. An example would be for a 10-item list like in this study, recollection of the item presented first would score a 10, the second word presented would score a 9, and this will continue until all words have been presented. The formula for scoring each item by this scoring measure is:

                RWR =          List Length      -     Presentation Position      +     1 (Buschke et al., 2006).                                                                                                                                               

The recency ratio on the other hand is calculated by dividing the recency scores in the immediate recall trial by the corresponding scores in the delayed recall trial. This then generated a recency (Rr) score. A correction was then applied to correct for missing data and zero scores. The correction looks like; (immediate recency score + 1)/ (delayed recency score + 1) (Bruno et al., 2018).



List learning data for was coded in a binary manner to denote whether each item in each trial was correctly recalled (0 = incorrectly or not recalled; 1 = correctly recalled). Words were assigned to primacy and recency based on the order in which they were recalled. If a word was recalled first it was assigned rank order 1, it was then checked to see in what section of the word list it was recalled. If the word was recalled in the first three words, this was assigned to the primacy group. If a word was recalled in the middle words (4, 5, 6 7) it was assigned to a neutral group with no bias. If the word recalled was in the final three words of the list is was in the recency group. Using these categories, it can be seen if there is a bias towards the primacy or recency effect


Statistical analysis

Analysis was performed by SPSS 25. Variables included in the main analysis were total and delayed recall and primacy and recency scores. Along with this there was the number of primacy and recency scores recalled after trial 1 of the list learning.

Each primacy and recency score were correlated with general cognitive ability using a Pearson’s correlation. A Mann-Whitney U test was used to determine the difference between the dementia and depression group difference between the dementia and depression group and participant’s education and general RBANS score (Total scale score).

The principal research question is examined by using a binary logistic regression with recency scores (recency ratio, recency words recalled after trial 1 and total recency words recalled. Using data from both immediate and delayed memory), Primacy scores (Primacy words recalled after trial 1 of list learning and total Primacy words recalled). The Binary logistic regression will be used separately for the primacy effect analysis and the recency effect analysis.


The sample used in this study (N = 33) was made up of a variety of ages, sex and educational background. The mean age of participants was M = 70.48 years old. There was an unequal balance between males (n = 14) and females (n = 19) with females having slightly more participants in total. The educational background of participants was widely different, with education ranging from primary school to university level education. The most common education found was primary school (n = 15), while the least frequent was university level education (n = 3). Group/inter certificate education was second most frequent education (n = 8) with leaving certificate educational level following closely behind it (n = 7). Frequency and descriptive statistical analysis were used to determine the difference in the scores of list learning total (M= 18.03, SD = 5.29) and story recall total (M= 5.30, SD= 3.17).  Further analysis was conducted to highlight this difference between diagnostic groups. For list learning dementia (M= 17.36, SD= 4.01) had lower mean scores for the total words recalled than did depression (M= 19.36, SD= 7.27). This is also the case for the story recall total with dementia patients (M= 4.59, SD= 3.32) seen in table 1, showing lower scores than those with depression (M= 6.73, SD= 2.37) which can be seen in table 2.


Table 1




List Learning Total Score

Story Recall Total







Std. Deviation







Table 2




List Learning Total Score

Story Recall Total






Std. Deviation






The serial position effect can also be seen for the list learning trial scores, with the Primacy words (= 220) and recency words (N = 226) being highly prevalent among the four list learning trials with the lowest words recalled being seen in the neutral position (N = 152), this presents as a U shaped curve, denoting the serial position effect and can be clearly seen in Figure 1.  






Figure 1




The same effect was seen when the diagnosis groups were split, and the scores analysed individually. Patients with dementia were examined, the words recalled in each position were totalled and the serial position effect is again seen in figure 2. The recency effect (N = 159) is slightly more prevalent than the primacy effect (N = 134) in patients with dementia.




Figure 2






Words recalled in each position were summed for patients with depression and found the opposite effect to patients with dementia, with the primacy effect (N = 86) showing a slight prevalence to the recency effect (N = 67) and the neutral section (N = 60). This is seen clearly in figure 3. Again, there is a distinguishable serial position curve.



Figure 3



 Moving on, correlational analysis was conducted to determine an association between general RBANS performance and age, to determine whether age is associated with performance differences. It was found that age had no significant relationship between any of the five tests of the RBANS measuring general cognitive performance; Immediate memory (r = -.11, p = .53), Visuospatial/Constructional (r = .21, p = .24), language (r = .16, p = .36), attention (r = -.08, p = .65), delayed memory (r = .14, p = .45) and total score (r = .04, p = .82) with the highest Pearson’s correlation being between age and visuospatial/constructional (r = 0.21, p .05). Moving on, a Pearson’s correlation was conducted to examine a relationship between diagnostic group and education. There was no significant correlation found between diagnosis and education (r = .23, p .19). From this homogeneity of variance and normality of distribution was tested for before further directional analysis was conducted. It was concluded that the data was not normally distributed (Shapiro-Wilks p = .00) and did not display homogeneity of variance (Levene Statistic = 5.68, p = .003). Consequently, a non-parametric test was used for further analysis. For further analysis, a Mann-Whitney U test was used. The test showed that there was not a significant relationship between education and diagnosis group = 98.5 (p = .36), meaning that age did was not associated with either condition. A Mann-Whitney U test was then conducted to observe the relationship between General RBANS performance (measured by the total scale number) and diagnosis. It was found that there was no significant relationship between general RBANS performance and diagnosis U = 70 (= .051). A binary logistic regression was used to examine a statistically significant relationship between diagnosis and several predictor variables. Initially, it was tested to see if there was a significant relationship between diagnosis and both the recency and primacy effect. The test was carried out and in the first block of the regression was just age. The second block contained the number of recency words recalled, the number of recency words for trial 1 and the recency ratio. There was no significant predictor for block 1 of age and diagnosis ( = 1.19, p = .27, Exp(B) = 2.5). For block two the recency ratio did not act as a predictor for diagnosis ( = 1.65, = .19, Exp(B) = .148, 95% CI = .008, 2.73). It was also found that recency words recalled in total for List learning shoed no relationship to diagnosis ( =2.3, = .13, Exp(B) = .462, 95% CI = .17, 1.3), while the recency words for trial 1 of List learning also showed no relationship to diagnosis ( = 1.59, = .20, Exp(B) = 5.66, 95% CI = ..383, 83.5) A second binary logistic regression was done to examine the relationship between diagnosis and the primacy statistics. Block 1 analysis was the same as before, but the second block of tests yielded a very similar statistical outcome. Block 2 of the primacy logistic regression showed no statistical significance between the number of primacy words recalled in trial 1 of list learning and diagnosis (= 3.2, p = .07, Exp(B) = 41.52, 95% CI = .71, 2425.91)  this is similar to the number of primacy words recalled in total for list learning also failing to produce a significant result (= .62, p = .43, Exp(B) = .68, 95% CI = .25, 1.8) this means that neither method used to measure the primacy effect shoed a relationship to diagnosis. It follows that a diagnosis of dementia or depression cannot be predicted using the serial position effect, the recency ratio/effect and the primacy effect.



The aim of this study was to determine a relationship between the recency ratio and a diagnosis of dementia and depression. Previous research has found that the recency ratio is a predictor of mild cognitive impairment and can also be seen as a predictor of dementia. Alzheimer’s disease is a form of dementia and it presents neuropathological degeneration in the hippocampal area and the surrounding medial temporal lobe (Raj et al., 2015). A consequence of this is poor episodic memory and in particular there is a difficulty forming new memories (Bruno, Reichert Pomara, 2016).  Memory deficits are markers for the disease also, with the recency effect being a prominent marker for Dementia when using the serial position effect of free recall. Patients with dementia tend to lose the items at the beginning of an immediately learned list (primacy effect), while their ability to acquire recently learned items remains intact, this is known as the recency effect (Bruno, Reichert Pomara, 2016)

Previous research has examined serial position effects have long been used to measure memory capabilities of individuals with Alzheimer’s disease (Carlesimo, Fadda, Sabaddini Caltagirone, 1996) and also for people with mild cognitive impairment, which h can manifest with cognitive impairments seen in people with depression (Howieson et al., 2010). There are several different measures attempting to calculate the serial position effect on these individuals with impairments, one notable measure that was used in this study is the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). This is a battery of tests measuring neuropsychological status and cognitive functioning. The reliability of the RBANS test has been supported and the clinical applications have been approved by literature from McKay, Casey, Wertheimer Fichtenberg (2007). These researchers have concluded that the RBANS test proved to be a useful measure for brief screening of dementia and Alzheimer’s disease. We chose this measure as research has supported this test to be more useful than other tests for screening brain injury (Carone et al., 2004)


Recency effect

The results of this study found that there was no statistically significant relationship between a diagnosis of dementia or depression and the recency effect.  Our proposed hypothesis was that patients with a diagnosis of dementia would exhibit a recency effect in their memory test results, while patients with a diagnosis of depression would exhibit a primacy effect in their memory tests. Our data did not support this hypothesis as the data exhibited no bias towards recency or primacy effect for patients with a diagnosis of dementia. Research by Burkart, Meun Benkert (1998) who found that in in patients with dementia showed an intact recency effect while a slightly impaired primacy effect was found. This is contrary to our data findings as there was no relationship found between the recency effect for patients with dementia. Oppositely to recent research, no relationship between immediate memory tasks, delayed memory tasks, list learning and story recall was found to be significantly associated with diagnosis of the individual.

The approach taken to calculating the recency ratio is one based of previous research by Bruno et al. (2018) where the scores were taken from the list learning task and the delayed recall task. Scores from the list learning task of the RBANS were taken from trial 1. The number of recency words recalled were divided by the number of recency words recalled in the delayed recall task. A correction was added to this in order to counteract zero scores. Recency words for the immediate and delayed recall tasks were the final three words in the 10-word list and these scores were measured by totalling the number of recency words recalled. Without a significant recency effect participant showed no difference in the position they recalled words. The failure to detect a significant recency effect is contrary to research in this area and this may be a result of the method of scoring used.



Primacy effect

Our results, similar to the recency effect, found no relationship between participants with depression and the primacy effect. As part of the proposed hypothesis of the study, it was stated that there would be a primacy effect for participants diagnosed with a form of mild cognitive impairment which often exhibits cognitive difficulties in conditions such as depression supported by research from Talamonti et al. (2019). This study found an effect for primacy in patients with mild cognitive impairment and this is then supported by Bruno et al. (2013) which shows that delayed primacy effect is the most sensitive predictor of mild cognitive impairment in older adults. In the study by Talamonti et al. (2019) they calculated primacy scores by totalling the number of primacy words recalled and dividing this number by the number of possible primacy words that could be recalled, this is different to the method used in our study as we did not divide the number of words recalled by the possible amount that could be recalled, rather we just totalled the recalled primacy words. Primacy words in our study also differed as we used the first three words of the 10-word list as our primacy words criteria. Our data did not show a relationship between the primacy effect and those diagnosed with depression. This, in part, could be a result of the small sample size employed. Small sample size may be a factor in our data not being significant because our data for the primacy effect narrowly missed out on significance level results. 


While small sample size is a limitation of this study, the sample size was small for the study but there was also inequality in the size of the diagnosis group with dementia (n = 22) having far more participants than the depression group (n = 11). This would affect the distribution of scores as well as potentially being a factor in the results being insignificant. Another limitation of the study would be the methodology used to calculate a recency and primacy variable. The method used was the recency ratio and the total number of primacy and recency words recalled for both trial 1 of list learning and the overall of immediate and delayed memory. Other recent research with significant findings used a different measure that divided the total number of words by the number of words that could possibly be recalled for a given effect. An example of this may be the total amount of recency words recalled would be divided by the number of recency words that could be recalled. This may act as a better variable for measurement of these effects. Going forward with research, an increased sample size would be sought after as the current one is too small. Other methods of measuring the primacy and recency effect would be used and inspiration taken from previous research that has yielded significant findings between these variables. With the implementation of these changes and as a result of previous significant research it is beneficial for this area of Psychology to continue research in the domain of dementia and depression and whether these conditions exhibit the recency or primacy effect respectively.