Theoretical and Clinical Basis of the FCSRT
The Free and Cued Selective Reminding Test (FCSRT) is grounded in the principle of encoding specificity, whereby retrieval is more successful when the cues available at recall correspond to the semantic operations established during learning.[1] Unlike conventional list-learning tests, in which poor recall can reflect weak attention, inefficient strategy use, inadequate semantic organization, retrieval failure, or impaired storage, the FCSRT constrains the encoding operation by requiring the participant to identify each item in response to a unique semantic category cue. This procedure provides direct evidence that the target item has been perceived, named, and linked to its retrieval cue before recall is tested.[2,3]
Controlled encoding reduces, but does not eliminate, the contribution of attentional and executive inefficiency to subsequent memory performance. Free recall then measures the capacity for self-initiated retrieval, whereas cued recall assesses whether information that was not produced spontaneously remains accessible when the original semantic cue is reinstated. Marked improvement with cueing is commonly interpreted as evidence that at least part of the difficulty lies in strategic retrieval. Conversely, persistently low total recall despite verified encoding and cueing is more consistent with impaired consolidation or storage. This contrast underlies the clinical concept of the amnestic syndrome of the hippocampal type.[3,4]
This interpretation is probabilistic rather than anatomically specific. Limited benefit from cueing is strongly associated with Alzheimer’s disease (AD) and medial temporal dysfunction, but it is not unique to AD; genuine memory impairment may also occur in other neurodegenerative disorders, including some cases of behavioral-variant frontotemporal dementia. Likewise, cueing can be ineffective when the category itself is not understood, language output is impaired, or task engagement is insufficient. The FCSRT must therefore be interpreted in relation to the broader cognitive profile, functional history, neurological findings, and, when available, biomarker and neuroimaging evidence.[5,6]
Episodic memory impairment in the early stages of AD is largely associated with damage to the episodic memory network encompassing the Papez circuit, starting with the hippocampus, and primarily affects memory storage.[7,8] Standard free-recall tests may be insufficient to distinguish this impairment from other memory disorder profiles, such as secondary memory difficulties arising from attentional or executive dysfunction, as poor performance may result from inadequate encoding, retrieval difficulty, or a genuine storage deficit.[2,3] The FCSRT was developed to reduce this ambiguity. Through a controlled learning phase, it ensures that each item is processed deeply and in a category-specific manner during encoding and then provides the same category cues during retrieval.[2,3]
This design enables the FCSRT to help distinguish a genuine storage deficit, in which poor free recall does not improve sufficiently with cueing, from a retrieval-based difficulty, in which poor free recall improves substantially with cueing.[3,7] This distinction is particularly important in clinical practice for the early identification of the hippocampal-type amnestic syndrome associated with AD, even at the stage of mild cognitive impairment (MCI).[4,9] Performance in FCSRT is also associated with hippocampal atrophy and may begin to decline before the onset of AD dementia.[8,10]
The 16-item version of the FCSRT has been adapted into numerous languages and cultures worldwide. A comprehensive comparison of these adaptations in clinical and research settings has been presented elsewhere[11] and is not repeated in full here. Three parallel Turkish forms (FCSRT-16TR) have been developed, and their psychometric characteristics and reference data have been reported in a separate study.[12] However, psychometric evidence alone does not ensure consistent clinical use; administration, scoring, and interpretation must also be standardized across clinicians. Therefore, this narrative review and clinical practice guide synthesizes the relevant evidence and provides a practical, step-by-step framework for use of the FCSRT-16TR in Türkiye.
The evidence base is substantial but heterogeneous. A recent systematic review identified 64 FCSRT-Immediate Recall (FCSRT-IR) studies, including normative, clinical validation, and biomarker investigations.[11] Overall findings supported the clinical utility of the FCSRT-IR in MCI and dementia, particularly within the AD spectrum. At the same time, differences in modality, population, scoring, and diagnostic purpose limited direct comparison across studies, reinforcing the need for version-specific administration and interpretation.
PROTOCOL VARIABILITY AND THE POSITION OF THE TURKISH FORMS
The label FCSRT does not refer to a single, completely uniform procedure. Published versions differ in stimulus modality (printed words versus pictures), item number, language-specific word lists, the presence or absence of an immediate-recall verification phase, the number and duration of learning trials, interference tasks, delay interval, recognition procedures, scoring variables, and the formula used to quantify sensitivity to cueing. Word and picture versions should not be considered interchangeable: performance is generally higher with pictures, potentially because they support both visual and verbal encoding.[13] Even apparently minor procedural changes can therefore alter task difficulty, ceiling effects, and the relative contribution of language, visual processing, and executive control.[11,13]
The FCSRT-16TR retains the core architecture of controlled semantic encoding followed by three free- and cued-recall trials and delayed free and total recall. Its distinctive features are culturally and linguistically selected Turkish target words and category cues, three parallel forms (A, B, and C) intended to support repeated assessment, and an extended recognition condition containing target words, semantically related lures, and unrelated lures. These elements were developed for Turkish-speaking examinees and should not be assumed to be psychometrically equivalent to a direct translation of another national version. Each form should be administered, scored, and interpreted within the corresponding Turkish protocol and reference framework.[12]
Parallel forms reduce repeated exposure to identical words but do not automatically guarantee complete score equivalence. When serial assessment is clinically necessary, the form used, interval between administrations, prior exposure, disease progression, and possible practice effects should be documented. Change across visits should be interpreted against evidence for equivalence and measurement error rather than assumed to represent true cognitive improvement or decline solely because different word lists were used.
TEST MATERIALS AND STRUCTURE
The FCSRT-16TR comprises 16 items presented on four cards, with four items on each card. Each item belongs to a distinct semantic category, and each word on a card is assigned to only one category. This arrangement ensures that the category cue identifies a single item during retrieval. Three parallel forms (A, B, and C) have been developed for clinical and research use. The forms contain different item sets to minimize practice effects during repeated administrations but were designed to be structurally equivalent.
The material set consists of (i) four A4-sized stimulus cards, each containing four items; (ii) a record form for documenting controlled learning, free recall, cued recall, delayed recall, and recognition responses; (iii) a recognition list comprising 16 target words, 16 semantically related distractors, and 16 unrelated distractors; and (iv) a stopwatch.”
To protect test security and minimize prior exposure to the stimuli, the complete target-word lists and their corresponding semantic category cues are not reproduced in this article and should not be included in publications or other publicly accessible materials. The standardized FCSRT-16TR materials may be requested from the corresponding author for appropriate clinical or research use.
PREADMINISTRATION PREPARATION
The test should be administered individually in a quiet environment free from distractions. The examiner should be thoroughly familiar with the materials and instructions before administration, as strict adherence to the standardized procedure is essential for valid interpretation of the data. Inconsistent wording of the instructions across sessions may influence the degree of improvement with cueing and thereby compromise clinical interpretation.
The patient’s hearing and visual acuity should be assessed to ensure suitability for testing, and corrective aids such as glasses or hearing devices should be used when necessary. The test should be administered to individuals whose native language is Turkish or who have sufficient proficiency in Turkish. Administration is not recommended in patients with aphasia.
CLINICAL POPULATIONS AND BOUNDARIES OF INTERPRETATION
The FCSRT-16TR is most informative in adults presenting with subjective memory complaints, MCI, or mild to moderate dementia when the clinical question concerns the nature of the memory deficit. It can contribute to the characterization of suspected typical AD and may assist comparison with retrieval-dominant profiles associated with executive or subcortical dysfunction. It can also be used as one component of a broader assessment in other neurodegenerative conditions, including frontotemporal, Lewy body, vascular, and movement disorders. However, overlap between disease groups is substantial, and no FCSRT profile is pathognomonic for a single etiology.[5,6]
Interpretation is particularly constrained when the prerequisites for valid semantic encoding are not met. In aphasia, naming impairment, auditory-verbal comprehension difficulty, or impaired speech output may lower identification and recall scores independently of episodic memory. Particularly, in semantic variant aphasia, also known as semantic dementia, the participant may not reliably understand the category cue or the relationship between the cue and target; failure to benefit from cueing can then mimic a storage deficit. The examiner should document comprehension errors, naming failures, semantic substitutions, and the number of repetitions required during encoding. When these problems are prominent, the test should not be used to make strong inferences about hippocampal storage.
Severe visual impairment can prevent reliable identification of printed stimuli, whereas hearing impairment can compromise comprehension of instructions and cues. Corrective devices, adequate illumination, and verification of stimulus perception are essential; modifications that depart from the standardized procedure must be documented because they may invalidate comparison with norms.
STEP-BY-STEP ADMINISTRATION PROCEDURE
The steps below follow the standard controlled learning-selective reminding paradigm of the FCSRT-16TR. The procedure is based on the work of Van der Linden et al.,[14] which used 16 verbal items.
1. Encoding phase
The examiner shows the first card to the patient and reads the complete instruction: “You can see four words here. Among these four words, there is a [category name]. Which one is a [category name]?” The patient is asked both to say the correct word aloud and to point to it. For the remaining three cards, the abbreviated instruction is used: “Which one is the [category name]?”
On each card, the categories are queried in a fixed order: upper left, upper right, lower left, and lower right. The purpose of this phase is to ensure that every word is associated with its corresponding category cue during encoding. Therefore, no category should be omitted, and the relevant category cue must be provided even if the patient spontaneously names the correct word. If the patient selects or names an incorrect item during the identification phase, the examiner provides the correct response and repeats the category-item association. If severe semantic or language impairment prevents the reliable establishment of the category-item associations, administration should be discontinued because the subsequent recall scores cannot be validly interpreted within the controlled-encoding framework.
After the patient has correctly identified all four words on a card, the card is covered. The examiner then asks the following question for each category in sequence: “Among the four words you just saw, there was a [category name]. What was it?” All four categories are queried individually, regardless of whether the patient recalls each word. Each word correctly recalled on the first query receives one point, and the sum is recorded as the Immediate Recall (IR) score.
After all four categories have been queried and the IR score has been determined, a recovery procedure is applied to words that were not recalled. The card is shown again, and only the category cue corresponding to the unrecalled word is provided to strengthen encoding: “Which one was the [category name]?” After the patient points to and says the word, the card is covered again and the examiner asks, “What was the [category name]?” This show-cover-query cycle is repeated for the unrecalled item, up to a maximum of three encoding attempts.
Only correct responses produced during the first query are included in the IR score. Repetitions during the recovery procedure are used solely to reinforce encoding and are not scored.
2. Learning phase
After all four cards have been completed, the cards are removed, and an interference task is administered. The examiner says, “Now I would like you to count backward by ones, starting from 374, until I ask you to stop.” The task is timed for 20 sec.
FREE- AND TOTAL-RECALL SCORES
“A short while ago, we learned a list of words together. Now I would like you to tell me all the words you remember from that list, in any order, as they come to mind.” The examiner records the words spontaneously recalled by the participant in the Free Recall column for the first trial, numbering them in the order in which they are produced. A maximum of 2 min is allowed for free recall.
For each word not recalled spontaneously, the corresponding category cue is provided: “There was a [category name] on the list. What was it?” During the first and second trials, if the participant does not produce the target word after the category cue, or produces an incorrect word from the same category, the examiner states the correct target word as a reminder.
The number of words recalled without cues constitutes the Free Recall score for that trial. The total number of words correctly retrieved through free and cued recall constitutes the Total Recall score for that trial.
Following the first trial, a 20-sec interference task is administered, during which the participant counts backward by ones from 294. The same procedure is repeated after the second trial, with the participant counting backward by ones from 188 for 20 sec.
The third trial differs from the first two in one important respect: if the participant fails to retrieve a target word after presentation of its category cue, the examiner does not provide the correct target word or any other corrective feedback. After the third trial, any target words that remain unrecalled are not re-presented or relearned before proceeding to the recognition and delayed-recall phases.
After completion of all three trials, the number of words freely recalled across the trials is summed to obtain the Total Free Recall (TFR) score. The number of words correctly retrieved through free and cued recall across all three trials is summed to obtain the Total Recall score.
RECORDING AND HANDLING OF ERROR RESPONSES
Intrusions, semantic substitutions, and repetitions should be recorded throughout the free- and cued-recall trials. During free recall, incorrect responses should be recorded but not corrected, and no feedback regarding their accuracy should be provided. The participant should simply be allowed to continue recalling. Repeated production of a previously recalled target should be recorded as a repetition and should not receive an additional point. Responses that were not part of the learned list should be recorded as intrusions and should not contribute to the Free Recall score.
During cued recall, incorrect responses, including semantic substitutions and other intrusions elicited by the category cue, should also be recorded. In the first and second trials, if the participant fails to retrieve the target word or produces an incorrect response following the category cue, the examiner provides the correct target word according to the corrective procedure described above. In the third trial, incorrect responses are recorded but no corrective feedback is provided.
3. Recognition
A total of 48 words are presented to the patient one at a time: the 16 target words from the learning list, 16 distractor words semantically related to the targets, and 16 semantically unrelated distractor words.
The examiner provides the following instruction: “Now I am going to show you some words. Some of these words were on the list we learned earlier, and some were not. For example, was [first word] on the list you learned earlier? What about [second word]? Was it on the list?” The same query format is applied sequentially to every word on the list.
Recognition is not implemented uniformly across FCSRT protocols. In a recent systematic review, only three normative studies using the word version included a recognition phase, while one additional study specifically developed normative data for a recognition task.[11] The defining components of the classical procedure remain controlled semantic encoding and free and cued recall. A Spanish recognition extension combined the 16 targets with semantically related and unrelated foils, illustrating that recognition measures require their own validation and normative framework.[15] The FCSRT-16TR recognition condition is likewise an extension specific to the Turkish protocol: its 16 semantically related and 16 unrelated lures permit separate examination of target hits and false-positive responses under different degrees of semantic competition. These findings should be interpreted alongside free and cued recall rather than as a substitute for these measures or as directly equivalent to recognition scores from other versions.
4. Delayed free and total recall
Following completion of the learning trials and the immediate recognition phase, a delay of 20 to 30 min is introduced. This interval should be filled with other neuropsychological tests that do not interfere with the FCSRT-16TR administration procedure or its words, preferably tests that use visual materials. At the end of the delay, the patient is asked to recall all the learned words without any cues. The number of words correctly recalled spontaneously is recorded as the Delayed Free Recall score.
The corresponding category cues are then provided for words that were not freely recalled. The total number of words correctly recalled through free and cued recall constitutes the Delayed Total Recall score.
Scoring
The principal scoring variables derived from the FCSRT-16TR are summarized in this section. Total Free Recall (TFR) is the sum of words correctly recalled without any cues across the three learning trials. The maximum score is 48. Total Recall is the total number of words correctly recalled through free recall and category cueing across the three learning trials. The maximum score is 48. Delayed Free Recall is the number of words correctly recalled without any cues after the delay. The maximum score is 16. Delayed Total Recall is the total number of words correctly recalled through free recall and category cueing after the delay. The maximum score is 16. Index of Sensitivity to Cueing (Cue Index, CI): A proportional measure of the extent to which category cues recover words that were not retrieved during free recall. It is calculated using the following formula:[3,16] CI = (Total Recall-TFR)/ (48-TFR).
A high index suggests that the information has been stored but is difficult to retrieve spontaneously and that the individual benefits from category cues. Low index values indicate limited benefit from cueing and support the possibility of a storage deficit.
Recognition is defined by the total number of correctly recognized words among the 16 target words learned previously. Each correctly recognized target receives one point, yielding a maximum score of 16. In addition, “old” responses to closely related and unrelated distractors are recorded separately as false recognitions.
Scores should be entered on the record form separately for every word in every trial. It is important to code which words were recalled freely, which were retrieved with a category cue, and which distractors elicited false-positive responses during recognition. This detailed recording permits retrospective error analyses and a more comprehensive evaluation of the clinical memory profile.
PRINCIPLES OF CLINICAL INTERPRETATION
Interpret the recall pattern, not an isolated score
The central clinical contribution of the FCSRT is the comparison between spontaneous retrieval and retrieval after reinstatement of the semantic cues used at encoding. Free recall reflects self-initiated retrieval under controlled learning conditions. Cued recall indicates whether items not produced spontaneously remain accessible when the original cue is supplied. Substantial improvement with cueing may be compatible with retrieval inefficiency, whereas persistently poor total recall despite successful controlled encoding and repeated cueing is more consistent with a storage impairment. This distinction is probabilistic and should be evaluated across learning trials and delayed recall rather than inferred from a single value.
The examiner should first verify that the participant perceived, named, and semantically linked every target during encoding. Errors caused by aphasia, impaired semantic knowledge, reduced auditory comprehension, sensory limitations, fatigue, or poor engagement can reduce apparent benefit from cueing and mimic a storage deficit. Intrusions, semantic substitutions, repetitions, response latency, need for repeated encoding, and changes across trials should therefore be considered alongside quantitative scores.
USE VERSION-SPECIFIC REFERENCE INFORMATION
Interpretation should use reference information developed for the same language, stimulus modality, word set, administration sequence, delay, and scoring definition. Performance is influenced by age, education, language, culture, and characteristics of the test version; values reported for picture-based or other national adaptations should not be transferred mechanically to the FCSRT-16TR. For reference values derived from cognitively healthy Turkish adults and guidance on demographic interpretation, see Dursun-Uncu et al.[12]
INTEGRATE THE TEST WITH THE BROADER CLINICAL ASSESSMENT
The FCSRT-16TR must not be used as a stand-alone diagnostic test for AD. Clinical interpretation requires integration of symptom history, functional change, premorbid ability, mood and medical factors, neurological findings, the wider neuropsychological profile, and, when indicated, neuroimaging and fluid or molecular biomarkers. Low free and total recall can increase suspicion of an amnestic syndrome, but similar impairment may arise in other neurological, psychiatric, developmental, or systemic conditions.[5,6]
Group-level studies link poorer free and total recall or reduced benefit from cueing with medial temporal or hippocampal atrophy, amyloid burden, tau-related pathology, and abnormal cerebrospinal fluid or blood biomarkers.[5] Free recall may change relatively early, while deterioration of total recall and cued performance may accompany broader involvement of the memory system.[11] These associations support the mechanistic interpretation of the test but do not permit an individual FCSRT result to substitute for biomarker assessment.
LONGITUDINAL AND PARALLEL-FORM INTERPRETATION
When the test is repeated, document the form administered, interval between assessments, previous exposure, clinical change, and factors that may affect practice. Parallel forms reduce exposure to identical targets but do not guarantee perfect score equivalence. Apparent change should be judged in light of the evidence for form equivalence, measurement error, and the complete clinical context rather than attributed automatically to cognitive improvement or decline.
Limitations and future directions
The FCSRT literature encompasses multiple procedures that share a conceptual foundation but differ in stimulus modality, learning structure, delay, recognition design, and scoring. This heterogeneity limits cross-study comparability and means that evidence obtained with one version may not generalize fully to another. The recognition extension of the FCSRT-16TR also requires further clinical validation across diagnostic groups.
Administration and interpretation may be limited in individuals with severe hearing or visual impairment, insufficient Turkish proficiency, marked language or semantic impairment, or advanced dementia. Future studies should evaluate multicenter feasibility, interrater scoring reliability, longitudinal sensitivity, parallel-form change, and the clinical contribution of the recognition condition. Digital administration and automated scoring may improve standardization if equivalence with the established protocol is demonstrated.
In conclusion, through controlled learning and category cueing, the FCSRT-16TR can help clinicians characterize whether memory difficulty is dominated by inefficient spontaneous retrieval or by impaired retention despite supported encoding and retrieval. Its clinical value depends on faithful administration, detailed error recording, version-specific interpretation, and integration with the broader assessment. This narrative review and practitioner's guide provides a shared procedural and interpretive framework for consistent use of the three Turkish forms.