Factor Analysis of Card Sort Data: An Alternative to Hierarchical Cluster Analysis

2005 ◽  
Vol 49 (5) ◽  
pp. 691-695 ◽  
Author(s):  
Miranda G. Capra
Keyword(s):  
Cluster Analysis ◽  
Factor Analysis ◽  
Statistical Analysis ◽  
Hierarchical Cluster Analysis ◽  
Hierarchical Cluster ◽  
Common Method ◽  
Card Sorting ◽  
Card Sort ◽  
Formal Statistical Analysis

Software and product designers use card sorting to understand item groups and relationships. In the usability community, a common method of formal statistical analysis for open card sort data is hierarchical cluster analysis, which results in a tree of the items sorted into distinct, nested clusters. Hierarchical cluster analysis is appropriate for highly structured settings, like software menus. However, many situations call for softer clusters, such as designing websites where multiple pages link to the same target page. Factor analysis summarizes the categories created in card sorts and generates clusters that can overlap. This paper explains how to prepare card sort data for statistical analysis, describes the results of factor analysis and how to interpret them, and discusses when hierarchical cluster analysis and factor analysis are appropriate.

scholarly journals Use of Factorial Analysis of Mixed Data (FAMD) and Hierarchical Cluster Analysis on Principal Component (HCPC) for Multivariate Analysis of Academic Performance of Industrial Engineering Programs

2020 ◽  
Vol 55 (5) ◽  
Author(s):  
Delimiro Visbal-Cadavid ◽  
Adel Mendoza-Mendoza ◽  
Enrique De La Hoz-Dominguez
Keyword(s):  
Higher Education ◽  
Cluster Analysis ◽  
Factor Analysis ◽  
Academic Performance ◽  
Hierarchical Cluster Analysis ◽  
Higher Education Institutions ◽  
Hierarchical Cluster ◽  
Industrial Engineering ◽  
Mixed Data ◽  
Engineering Programs

The article describes a new idea about using Factor Analysis, Mixed Data, and Hierarchical Cluster Analysis on Principal Components to study the academic performance in 82 Industrial Engineering Programs in Colombia. For this, we used the data from the results of the standardized test of the Saber Pro.). The authors find that the first three components explain 89.12% of the original data set variability. The quantitative variables associated with the Factor Analysis are the first dimension, while the two qualitative variables are related to the second dimension. The first factor explains 95.83% of the dispersion of the scores in Critical Reading, 94.72% of the variability in Quantitative Reasoning, 94.51% of the variation in Mathematics and Statistics, among others. This study shows a strong positive correlation between the quantitative variables and the first factorial axis. It assumes that the Industrial Engineering Programs of public higher education institutions perform better than private ones. The article stipulates that the higher education institutions belonging to the Andean Region present a better performance, followed by the higher education institutions located in the Pacific Region. In general terms, the results confirm that the best performing universities usually appear in the first places in the different rankings and are located in the big cities.

scholarly journals The organization of taste sensibilities in hamster chorda tympani nerve fibers.

1988 ◽  
Vol 91 (6) ◽  
pp. 861-896 ◽  
Author(s):  
M E Frank ◽  
S L Bieber ◽  
D V Smith
Keyword(s):  
Cluster Analysis ◽  
Factor Analysis ◽  
Hierarchical Cluster Analysis ◽  
Nerve Impulse ◽  
Hierarchical Cluster ◽  
Nerve Fibers ◽  
Chorda Tympani ◽  
Chorda Tympani Nerve ◽  
Activation Patterns ◽  
Response Profiles

Electrophysiological measurements of nerve impulse frequencies were used to explore the organization of taste sensibilities in single fibers of the hamster chorda tympani nerve. Moderately intense taste solutions that are either very similar or easily discriminated were applied to the anterior lingual surface. 40 response profiles or 13 stimulus activation patterns were considered variables and examined with multivariate statistical techniques. Three kinds of response profiles were seen in fibers that varied in their overall sensitivity to taste solutions. One profile (S) showed selectivity for sweeteners, a second (N) showed selectivity for sodium salts, and a third (H) showed sensitivity to salts, acids, and other compounds. Hierarchical cluster analysis indicated that profiles fell into discrete classes. Responses to many pairs of effective stimuli were covariant across profiles within a class, but some acidic stimuli had more idiosyncratic effects. Factor analysis of profiles identified two common factors, accounting for 77% of the variance. A unipolar factor was identified with the N profile, and a bipolar factor was identified with the S profile and its opposite, the H profile. Three stimulus activation patterns were elicited by taste solutions that varied in intensity of effect. Hierarchical cluster analysis indicated that the patterns fell into discrete classes. Factor analysis of patterns identified three common unipolar factors accounting for 82% of the variance. Eight stimuli (MgSO4, NH4Cl, KCl, citric acid, acetic acid, urea, quinine HCl, HCl) selectively activated fibers with H profiles, three stimuli (fructose, Na saccharin, sucrose) selectively activated fibers with S profiles, and two stimuli (NaNO3, NaCl) activated fibers with N profiles more strongly than fibers with H profiles. Stimuli that evoke different patterns taste distinct to hamsters. Stimuli that evoke the same pattern taste more similar. It was concluded that the hundreds of peripheral taste neurons that innervate the anterior tongue play one of three functional roles, providing information about one of three features that are shared by different chemical solutions.

scholarly journals Numerals in authorial Turkish-language texts and the stylometric analysis

2021 ◽  
Vol 270 ◽  
pp. 01038
Author(s):  
Andrei Zenkov ◽  
Eugene Zenkov ◽  
Miroslav Zenkov ◽  
Larisa Sazanova
Keyword(s):  
Cluster Analysis ◽  
Discourse Analysis ◽  
Statistical Analysis ◽  
Frequency Distribution ◽  
Computer Analysis ◽  
Hierarchical Cluster Analysis ◽  
Hierarchical Cluster ◽  
Literary Texts ◽  
Frequency Distributions ◽  
Turkish Language

Two approaches to the statistical analysis of texts are suggested, both based on the study of numerals occurrence in coherent texts. The first approach is related to the study of the frequency distribution of various leading digits of numerals occurring in the text. These frequencies are unequal: the digit 1 is strongly dominating; usually, the incidence of subsequent digits is monotonically decreasing. The frequencies of occurrence of the digit 1, as well as, to a lesser extent, the digits 2 and 3, are usually a characteristic author’s style feature, manifested in all (sufficiently long) texts of any author. This approach is convenient for testing whether a group of texts has common authorship: the latter is dubious if the frequency distributions are sufficiently different. The second approach is the extension of the first one and requires the study of the frequency distribution of numerals themselves (not their leading digits). The approach yields non-trivial information about the author, stylistic and genre peculiarities of the texts and is suited for the advanced discourse analysis. This paper deals with the application of the second approach to the literary texts in Turkish. We have analysed almost the whole corpus of works by are illustrated by examples of computer analysis of the literary texts by O. Pamuk and Y. Kemal – two of Turkey’s most prominent novelists. The hierarchical cluster analysis based on the occurrence of numerals in the texts by Pamuk and Kemal shows the author, genre, and chronology differences of numerals usage in the literary texts of these authors.

A corpus-based analysis of the verb pleróo in Ancient Greek

2017 ◽  
Vol 15 (1) ◽  
pp. 253-287 ◽  
Author(s):  
Georgios Ioannou
Keyword(s):  
Cluster Analysis ◽  
Statistical Analysis ◽  
Hierarchical Cluster Analysis ◽  
Multiple Correspondence Analysis ◽  
Hierarchical Cluster ◽  
Semantic Features ◽  
Ancient Greek ◽  
Image Schemas ◽  
Inductive Methods ◽  
Image Schema

Abstract This is a corpus-based study of the development of the verb pleróo in Ancient Greek, originally meaning fill, from the 6th c. bce in Classical Greek, up to the end of the 3rd c. bce in Hellenistic Koiné. It implements a hierarchical cluster analysis and a multiple correspondence analysis of the sum of the attested instances of pleróo of that period, divided by century. It explores the gains following a syncretism between two methodological strands: earlier introspective analyses postulating variant construals over intuitively grasped schematic configurations such as image schemas, and strictly inductive methods based on statistical analyses of correlations between co-occurring formal and semantic features. Thus, it examines the relevance of the container image-schema to the architecture of the schematic construction corresponding to the prototypical and historically preceding sense of pleróo, fill. Consequently, it observes how shifts in the featural configurations detected through statistical analysis, leading to the emergence of new senses, correspond to successive shifts on the perspectival salience of elements in the schematic construction of the verb.

Statistical Analysis of Parieto-Frontal Cognitive-Motor Networks

2009 ◽  
Vol 102 (3) ◽  
pp. 1911-1920 ◽  
Author(s):  
Bruno B. Averbeck ◽  
Alexandra Battaglia-Mayer ◽  
Carla Guglielmo ◽  
Roberto Caminiti
Keyword(s):  
Cluster Analysis ◽  
Statistical Analysis ◽  
Hierarchical Cluster Analysis ◽  
Hierarchical Cluster ◽  
Anatomical Connectivity ◽  
Statistical Structure ◽  
Cortical Areas ◽  
Prefrontal Cortical ◽  
Primate Brain ◽  
Statistical Regularities

Considerable information has been gathered on the anatomical connectivity within the parieto-frontal network of the primate brain. To examine the statistical regularities in this connectivity, we carried out hierarchical cluster analysis and found statistically significant clusters of areas: four in the parietal and six in the frontal lobe. Clusters were based on patterns of inputs from all cortical areas. Both parietal and frontal clusters were composed of sets of spatially contiguous architectonic areas. The four parietal clusters were composed of sets of anterior (somatosensory), dorsal, inferior, and medio-lateral parietal cortical areas. The six frontal clusters were composed of sets of dorsal premotor, ventral premotor, primary motor, cingulate motor, and dorsal and ventral prefrontal cortical areas. Furthermore, connectivity between frontal and parietal clusters was topographic and reciprocal. Thus we found substantial statistical structure and organization in the parieto-frontal network that gives a simplified but accurate description of this system.

Individual differences in wild (WWCPS) rat — manifested in the exploration box

2010 ◽  
Vol 41 (1) ◽  
pp. 31-35
Author(s):  
Wojciech Pisula
Keyword(s):  
Cluster Analysis ◽  
Factor Analysis ◽  
Individual Differences ◽  
Hierarchical Cluster Analysis ◽  
Hierarchical Cluster ◽  
Dimensional Structure ◽  
Two Dimensional ◽  
Factor Scores ◽  
Laboratory Rats ◽  
Two Factors

Individual differences in wild (WWCPS) rat — manifested in the exploration box Thirty nine WWCPS rats were tested in the exploration box throughout fifteen sessions. Factor analysis was run to extract the main dimensions describing rat behavior. Two factors were extracted, confirming the validity of the concept of two dimensional structure of individual differences in rats. Hierarchical cluster analysis run on factor scores showed that only three out of a possible four types of factor combinations are actually present within observed group of animals. In terms of individual differences structure, the study provide support for the view that laboratory rats are still rats.

Situating Valency Theory in functional-cognitive space

2018 ◽  
Vol 16 (2) ◽  
pp. 348-398
Author(s):  
Francisco Gonzálvez-García ◽  
Christopher S. Butler
Keyword(s):  
Cluster Analysis ◽  
Statistical Analysis ◽  
Multidimensional Scaling ◽  
Hierarchical Cluster Analysis ◽  
Hierarchical Cluster ◽  
Negative Evaluation ◽  
Questionnaire Item ◽  
Broad Group ◽  
Cognitive Space ◽  
German Syntax

Abstract This article builds on the work reported in Butler and Gonzálvez-García (2014), in which 16 functional and/or cognitive/constructionist theories were compared on the basis of questionnaires completed by experts and a reading of the literature on each approach. The aim is to extend this work to cover Valency Theory (VT henceforth), arguably the most widely used approach to the study of German syntax. We first report on a statistical analysis (correlation, multidimensional scaling and hierarchical cluster analysis) of the data from the questionnaires completed by two VT experts, in relation to those completed by experts in other approaches. We then present an analysis of each item in the questionnaire in relation to VT, leading to a positive or negative evaluation for each questionnaire item. The results are again analysed statistically. The picture that emerges is of a theory which, though distinctive, has clear relationships with a broad group of cognitively-oriented approaches.

scholarly journals Pengelompokan 33 Aksesi Kakao Berdasarkan Karakter Morfologi Komponen Buah

2017 ◽  
Vol 4 (2) ◽  
pp. 67
Author(s):  
Edi Wardiana ◽  
Juniaty Towaha ◽  
Syafaruddin Syafaruddin
Keyword(s):  
Cluster Analysis ◽  
Factor Analysis ◽  
Discriminant Analysis ◽  
Hierarchical Cluster Analysis ◽  
Morphological Character ◽  
Hierarchical Cluster ◽  
Dry Weight ◽  
Morphological Characters ◽  
Fresh Weight ◽  
Experimental Station

<em>Identification and grouping of diverse cacao accessions based on morphological character of pod components is important in breeding activities to generate superior varieties. This study aimed to grouping 33 cocoa accessions based on morphological characters of pod components, conducted at Pakuwon Experimental Station, Sukabumi, West Java, at an altitude of 450 asl with Latosol soil and B type climate (Schmidt and Fergusson), from January until December 2015. Observation was taken on 33 cacao accessions of Kaliwining (KW), planted in 2012 with 3 m x 3 m spacing under 26 years old Genjah Salak coconut trees as shading. Ten mature pods were randomly harvested from 10 cocoa plants of each accession at respective two seasons (February and December 2015). The variables observed were 7 characters of pod components i.e.: (1) fresh weight of pod, (2) number of beans, (3) fresh weight of beans, (4) dry weight of beans, (5) fresh weight of pod husk, (6) number of pod furrow, and (7) fresh weight of pulp. Data were analyzed by factor analysis followed by hierarchical cluster analysis in Ward's method and discriminant analysis. The result showed that 19 accessions were classified as high in bean and pod husk component characters, 9 accessions were classified as high in pulp characters, and 5 accessions i.e. KW 162, KW 528, KW 570, KW 571, and KW 720 were classified as high in all characters of pod components, and potential as parents in generating superior cacao varieties.</em>

scholarly journals Ανάλυση περιβαλλοντικού κινδύνου από βαρέα μέταλλα γεωγενούς και ανθρωπογενούς προέλευσης σε δύο αντιπροσωπευτικές περιοχές του Ελλαδικού χώρου

2016 ◽  
Author(s):  
Δήμητρα Μεγαδούκα
Keyword(s):  
Principal Component Analysis ◽  
Cluster Analysis ◽  
Factor Analysis ◽  
Hierarchical Cluster Analysis ◽  
Principal Component ◽  
Hierarchical Cluster ◽  
Component Analysis ◽  
Η Factor

Σκοπός της παρούσας μελέτης είναι η έρευνα έξι βαρέων μετάλλων σε δύο περιοχές του Ελλαδικού χώρου. Οι κεντρικοί άξονες της έρευνας αυτής αποτελούνται από τους εξής στόχους: i) να προσδιοριστούν οι συγκεντρώσεις των μετάλλων Pb, Cd, Zn, Co, Ni και Cr που εντοπίζονται σε αδιατάρακτα εδάφη, ii) να προσδιοριστούν οι πηγές προέλευσης των μετάλλων (ανθρωπογενής – γεωγενής) με εργαστηριακές και στατιστικές μεθόδους, iii) να προσδιοριστούν οι γεωχημικές μορφές των μετάλλων που κατανέμονται στα εδάφη, iv) να παρουσιαστούν γεωγραφικοί χάρτες κατανομής των μετάλλων με την χρήση του προγράμματος Γεωγραφικών Πληροφοριακών Συστημάτων, v) να προσδιοριστούν οι υποβόσκουσες σχέσεις μεταξύ των μετάλλων vi) να γίνει εκτίμηση περιβαλλοντικού κινδύνου που σχετίζεται με την ρύπανση εδαφών από βαρέα μέταλλα σε περιοχές επιβαρυμένες από ανθρωπογενή και γεωγενή αίτια.Επιλέχθηκαν δύο περιοχές μελέτης. α) Η Βόρεια Εύβοια επιλέχθηκε ως περίπτωση επιβάρυνσης από βαρέα μέταλλα γεωγενούς προέλευσης λόγω μητρικού υλικού - των βασικών και υπερβασικών πετρωμάτων που χαρακτηρίζουν γεωλογικά την περιοχή. β) Η Λαυρεωτική χερσόνησος επιλέχθηκε ως περίπτωση επιβάρυνσης από βαρέα μέταλλα ανθρωπογενούς προέλευσης από τις παλαιότερες (π.Χ.) αλλά και πρόσφατες μεταλλευτικές και μεταλλουργικές δραστηριότητες.Στην περιοχή της Βόρειας Εύβοιας επιλέχθηκαν 21 θέσεις δειγματοληψίας και στην Λαυρεωτική χερσόνησο επιλέχθηκαν 45 θέσεις δειγματοληψίας. Από κάθε θέση δειγματοληψίας, συλλέχθηκαν δείγματα από δύο διαφορετικά βάθη, 0-10 cm (Α) και 10-30 cm (Β). Σε όλα τα εδαφικά δείγματα προσδιορίστηκαν οι εξής φυσικοχημικές ιδιότητες των εδαφών : η κοκκομετρική σύσταση, το pH, το Eh, η οργανική ουσία και το ισοδύναμο CaCO3. Επίσης προσδιοριστήκαν οι μορφές των μετάλλων με την μέθοδο διαδοχικών εκχυλίσεων τροποποιημένη BCR. Η μέθοδος των διαδοχικών εκχυλίσεων στοχεύει επιλεκτικά και απελευθερώνει τα μέταλλα που είναι συνδεδεμένα σε : α) υδατοδιαλυτές & συνδεδεμένες με ανθρακικά άλατα, β) οξείδια Fe και Mn, γ) οργανική ουσία και δ) πυριτικές μορφές. Επίσης, για να γίνει μία σωστή και αντικειμενική σύγκριση των αποτελεσμάτων και να προσδιοριστούν οι σχέσεις μεταξύ των μετάλλων χρησιμοποιήθηκε το πρόγραμμα Factor Analysis (ανάλυση παραγόντων) και το πρόγραμμα Hierarchical Cluster Analysis. Η Factor Analysis με την μέθοδο της Principal Component Analysis, προσδιορίζει έναν μικρό αριθμό παραγόντων που εξηγούν το μεγαλύτερο μέρος της διακύμανσης που παρατηρείται, σε έναν πολύ μεγαλύτερο όγκο δεδομένων, ενώ η Hierarchical Cluster Analysis προσδιορίζει τις μεταβλητές ή τους παράγοντες, οι οποίοι εξηγούν τον τρόπο που συσχετίζονται τα μέταλλα απεικονιζόμενες σε ένα δενδρόγραμμα.

Sign in / Sign up

Export Citation Format

Share Document