scholarly journals Missing values in exploratory factor analysis: A ‘best of all possible worlds’ approach to imputation for incomplete survey data

2021 ◽  
Author(s):  
Bas Bosma ◽  
Arjen van Witteloostuijn
Keyword(s):  
Social Sciences ◽  
Factor Analysis ◽  
Exploratory Factor Analysis ◽  
Factor Structure ◽  
Possible Worlds ◽  
Missing Values ◽  
Ease Of Use ◽  
Factor Analyses ◽  
Imputation Methods ◽  
The Social

In the social sciences, multi-item scales and factor analyses are standard tools in survey research. In the social sciences, such tools are omnipresent, as are, unavoidably, nonresponses. The question is how to handle missing values when an exploratory factor analysis is intended. Deletion methods will result in — oftentimes substantial and damaging — reduction of power. The seemingly obvious alternative is to keep all respondents and apply imputation to missing values. However, with the true factor structure unknown, theoretically recommendable multiple imputation methods cannot simply be applied. Instead of declaring an entire method unsuitable for exploratory analysis, we propose an approach that keeps the relevant aspects of various methods and combines these by sacrificing less relevant aspects. Doing so, we keep understanding and ease of use in mind, aiming for an approach that is more rigorous and ‘correct’ than what is commonly used in practice, whilst still being straightforward enough to actually be used.

scholarly journals Missing values in exploratory factor analysis: A ‘best of all possible worlds’ approach to imputation for incomplete survey data

2021 ◽  
Author(s):  
Bas Bosma ◽  
Arjen van Witteloostuijn
Keyword(s):  
Social Sciences ◽  
Factor Analysis ◽  
Exploratory Factor Analysis ◽  
Factor Structure ◽  
Possible Worlds ◽  
Missing Values ◽  
Ease Of Use ◽  
Factor Analyses ◽  
Imputation Methods ◽  
The Social

In the social sciences, multi-item scales and factor analyses are standard tools in survey research. In the social sciences, such tools are omnipresent, as are, unavoidably, nonresponses. The question is how to handle missing values when an exploratory factor analysis is intended. Deletion methods will result in — oftentimes substantial and damaging — reduction of power. The seemingly obvious alternative is to keep all respondents and apply imputation to missing values. However, with the true factor structure unknown, theoretically recommendable multiple imputation methods cannot simply be applied. Instead of declaring an entire method unsuitable for exploratory analysis, we propose an approach that keeps the relevant aspects of various methods and combines these by sacrificing less relevant aspects. Doing so, we keep understanding and ease of use in mind, aiming for an approach that is more rigorous and ‘correct’ than what is commonly used in practice, whilst still being straightforward enough to actually be used.

scholarly journals Missing values in exploratory factor analysis: A ‘best of all possible worlds’ approach to imputation for incomplete survey data

2021 ◽  
Author(s):  
Bas Bosma ◽  
Arjen van Witteloostuijn
Keyword(s):  
Social Sciences ◽  
Factor Analysis ◽  
Exploratory Factor Analysis ◽  
Factor Structure ◽  
Possible Worlds ◽  
Missing Values ◽  
Ease Of Use ◽  
Factor Analyses ◽  
Imputation Methods ◽  
The Social

In the social sciences, multi-item scales and factor analyses are standard tools in survey research. In the social sciences, such tools are omnipresent, as are, unavoidably, nonresponses. The question is how to handle missing values when an exploratory factor analysis is intended. Deletion methods will result in — oftentimes substantial and damaging — reduction of power. The seemingly obvious alternative is to keep all respondents and apply imputation to missing values. However, with the true factor structure unknown, theoretically recommendable multiple imputation methods cannot simply be applied. Instead of declaring an entire method unsuitable for exploratory analysis, we propose an approach that keeps the relevant aspects of various methods and combines these by sacrificing less relevant aspects. Doing so, we keep understanding and ease of use in mind, aiming for an approach that is more rigorous and ‘correct’ than what is commonly used in practice, whilst still being straightforward enough to actually be used.

Dimensionality of burden in Alzheimer caregivers: confirmatory factor analysis and correlates of the Zarit Burden interview

2014 ◽  
Vol 26 (9) ◽  
pp. 1455-1463 ◽  
Author(s):  
Sheung-Tak Cheng ◽  
Timothy Kwok ◽  
Linda C. W. Lam
Keyword(s):  
Factor Analysis ◽  
Confirmatory Factor Analysis ◽  
Exploratory Factor Analysis ◽  
Factor Structure ◽  
Caregiver Burden ◽  
Care Recipient ◽  
Factor Analyses ◽  
Confirmatory Factor Analyses ◽  
Zarit Burden Interview ◽  
Confirmatory Factor

ABSTRACTBackground:To investigate dimensions of caregiver burden through factor analysis of the Zarit Burden Interview (ZBI), and to examine predictors of different dimensions of burden.Methods:Confirmatory factor analyses were performed on 395 Hong Kong Chinese Alzheimer caregivers to examine whether several proposed factor structures fit the data well. Subsequently, participants were split into two roughly equal subsamples, for the purpose of identifying the most optimal factor structure through exploratory factor analysis in Sample A (n = 183) and an independent verification through confirmatory factor analysis in Sample B (n = 212). ZBI subscales representing the established factors were correlated with caregiver and care-recipient variables known to be associated with burden.Results:Confirmatory factor analyses showed that factor models reported elsewhere did not fit the data well. Subsequently, exploratory factor analysis in Sample A suggested a 4-factor structure. After dropping three items due to poor factor loadings, the 4-factor structure was found to fit the data moderately well in Sample B. The four factors tapped personal strain, captivity, self-criticism, and loss of control. However, self-criticism was basically unrelated to the other three factors and showed a rather different pattern of correlations with caregiver and care-recipient variables. Self-criticism was more common among child caregivers and those who did not live with the care-recipient and was less involved in day-to-day care, yet feeling obligated and close to the care-recipient.Conclusions:The dimensions of caregiver burden may be culturally specific. More research is needed to examine cultural considerations in measuring caregiver burden.

Exploratory Factor Analysis of the French Version of the Big Five Questionnaire for Children (BFQ-C)

2017 ◽  
Vol 76 (3) ◽  
pp. 125-130 ◽  
Author(s):  
Martine Bouvard ◽  
Jean-Luc Roulin
Keyword(s):  
Factor Analysis ◽  
Exploratory Factor Analysis ◽  
Factor Structure ◽  
Big Five ◽  
Goodness Of Fit ◽  
Internal Validity ◽  
Factor Analyses ◽  
Confirmatory Factor Analyses ◽  
French Version ◽  
Exploratory Factor Analyses

Abstract. This article examines the internal validity of the French version of the Big Five Questionnaire for Children (BFQ-C). We first performed confirmatory factor analyses (CFA) to determine the fit of the factor structure identified in previous research on the BFQ-C and then used exploratory factor analyses. A sample of 399 children (192 boys and 207 girls) recruited from elementary schools completed the BFQ-C in their classrooms. Participants were 8 to 12 years old with a mean age of 9.58 years (SD = 0.98). The results indicated poor goodness-of-fit statistics for the CFA solution. Exploratory factor analysis improved the model fit markedly over conventional CFA. The factor structure of the French version of the BFQ-C suggested reasonable fit for the five intercorrelated factors corresponding to the Big Five with the exploratory factor analyses, even if the fifth factor appeared to be more problematic than the other 4 factors.

Relating Multidimensional Sets of Variables: Canonical Correlation or Factor Analysis?

1975 ◽  
Vol 36 (3) ◽  
pp. 859-862 ◽  
Author(s):  
John D. Morris ◽  
Wilson H. Guertin
Keyword(s):  
Social Sciences ◽  
Factor Analysis ◽  
Canonical Correlation ◽  
Common Factor ◽  
Research Problem ◽  
Canonical Analysis ◽  
Factor Analyses ◽  
Specific Research ◽  
The Social ◽  
Separate Factor

Several prominent multivariate psychological methodologists recommend the use of canonical correlation to relate multidimensional sets of variables. This method along with separate factor analyses of the sets is considered in relation to the questions they may be able to answer on a specific research problem. Alternate analyses of data from the social sciences illustrate the value of common factor analysis compared with canonical analysis as a method for relating the underlying constructs across sets of variables.

scholarly journals Development of a revised Jalowiec Coping Scale for use by emergency clinicians: a cross-sectional scale development study

BMJ Open ◽  
2019 ◽  
Vol 9 (12) ◽  
pp. e033053
Author(s):  
Jaimi H Greenslade ◽  
Marianne C Wallis ◽  
Amy Johnston ◽  
Eric Carlström ◽  
Daniel Wilhelms ◽  
...  
Keyword(s):  
Factor Analysis ◽  
Coping Strategies ◽  
Exploratory Factor Analysis ◽  
Factor Structure ◽  
Factor Analyses ◽  
Cross Sectional ◽  
Item Reduction ◽  
Urban Hospitals ◽  
Confirmatory Factor ◽  
Coping Scale

ObjectivesThe aim of this study was to develop and validate a scale to measure the coping strategies used by emergency staff in response to workplace stress. To achieve this aim, we developed a refined Jalowiec Coping Scale (JCS), termed the Jalowiec Coping Scale-Emergency Department (JCS-ED) and validated this scale on a sample of emergency clinicians.DesignA cross-sectional survey incorporating the JCS, the working environment scale-10 and a measure of workplace stressors was administered between July 2016 and June 2017. The JCS-ED was developed in three stages: 1) item reduction through content matter experts, 2) exploratory factor analysis for further item reduction and to identify the factor structure of the revised scale and 3) confirmatory factor analyses to confirm the factors identified within the exploratory factor analysis.SettingSix Emergency Departments (EDs) in Australia and four in Sweden. There were three tertiary hospitals, five large urban hospitals and two small urban hospitals.ParticipantsParticipants were eligible for inclusion if they worked full-time or part-time as medical or nursing staff in the study EDs. The median age of participants was 35 years (IQR: 28–45 years) and they had been working in the ED for a median of 5 years (IQR: 2–10 years). 79% were females and 76% were nurses.ResultsA total of 875 ED staff completed the survey (response rate 51%). The content matter experts reduced the 60-item scale to 32 items. Exploratory factor analyses then further reduced the scale to 18 items assessing three categories of coping: problem-focussed coping, positive emotion-focussed coping and negative emotion-focussed coping. Confirmatory factor analysis supported this three-factor structure. Negative coping strategies were associated with poor perceptions of the work environment and higher ratings of stress.ConclusionsThe JCS-ED assesses maladaptive coping strategies along with problem-focussed and emotion-focussed coping styles. It is a short instrument that is likely to be useful in measuring the types of coping strategies employed by staff.

scholarly journals The Evaluation of the Religious Commitment Inventory for Bulgarian Pomak Households

2020 ◽  
Vol 15 (6) ◽  
pp. 118
Author(s):  
Nikolaos Satsios ◽  
Kostas Karamanis ◽  
Aikaterini Galanou ◽  
Ioannis Sotiropoulos
Keyword(s):  
Social Sciences ◽  
Factor Analysis ◽  
Factor Structure ◽  
Religious Commitment ◽  
Interesting Aspect ◽  
Christian Community ◽  
The Social ◽  
Socio Demographic Factors ◽  
Consumer Marketing ◽  
The Impact

The aim of the research is to evaluate the Religious Commitment Inventory (RCI-10) on a sample of 400 Bulgarian Pomak households across 5 cities. For this purpose, factor analysis was used in order to confirm the factor structure of the religiosity scale. An interesting aspect assumes that the impact of religiosity on the consumer marketing as well as on the social sciences is especially important. Most significantly, this research demonstrates the universality of the religiosity correlation with socio-demographic factors, such as the place of residence, demonstrating the importance of the religiosity level. The results revealed higher levels of religiosity score in cities populated mainly by Muslims, while lower levels score were observed in cities where a Christian community was also present.

An investigation into the factor structure of the Cognitive Therapy Scale – Revised (CTS-R) in a CBT training sample

2021 ◽  
pp. 1-11
Author(s):  
Sarah Beale ◽  
Silia Vitoratou ◽  
Sheena Liness
Keyword(s):  
Factor Analysis ◽  
Confirmatory Factor Analysis ◽  
Cognitive Therapy ◽  
Exploratory Factor Analysis ◽  
Factor Structure ◽  
Factor Model ◽  
Single Factor ◽  
Behaviour Therapy ◽  
Cognitive Behaviour ◽  
Confirmatory Factor

Abstract Background: Effective monitoring of cognitive behaviour therapy (CBT) competence depends on psychometrically robust assessment methods. While the UK Cognitive Therapy Scale – Revised (CTS-R; Blackburn et al., 2001) has become a widely used competence measure in CBT training, practice and research, its underlying factor structure has never been investigated. Aims: This study aimed to present the first investigation into the factor structure of the CTS-R based on a large sample of postgraduate CBT trainee recordings. Method: Trainees (n = 382) provided 746 mid-treatment audio recordings for depression (n = 373) and anxiety (n = 373) cases scored on the CTS-R by expert markers. Tapes were split into two equal samples counterbalanced by diagnosis and with one tape per trainee. Exploratory factor analysis was conducted. The suggested factor structure and a widely used theoretical two-factor model were tested with confirmatory factor analysis. Measurement invariance was assessed by diagnostic group (depression versus anxiety). Results: Exploratory factor analysis suggested a single-factor solution (98.68% explained variance), which was supported by confirmatory factor analysis. All 12 CTS-R items were found to contribute to this single factor. The univariate model demonstrated full metric invariance and partial scalar invariance by diagnosis, with one item (item 10 – Conceptual Integration) demonstrating scalar non-invariance. Conclusions: Findings indicate that the CTS-R is a robust homogenous measure and do not support division into the widely used theoretical generic versus CBT-specific competency subscales. Investigation into the CTS-R factor structure in other populations is warranted.

Factor Structure of the K-ABC for Cantonese, English, and Punjabi Speaking Canadians

1988 ◽  
Vol 4 (1) ◽  
pp. 49-59
Author(s):  
Joanne M. Gardner
Keyword(s):  
Factor Analysis ◽  
Theoretical Model ◽  
Confirmatory Factor Analysis ◽  
Construct Validity ◽  
Internal Structure ◽  
Exploratory Factor Analysis ◽  
Factor Structure ◽  
Assessment Battery ◽  
Confirmatory Factor

The purpose of this study was to examine the construct validity (internal structure) of the Kaufman Assessment Battery for Children (K-ABC) for use with Cantonese, English, and Punjabi speaking Canadians. A confirmatory factor analysis revealed that the sequential/simultaneous theoretical model was supported by the English and Punjabi data: however, the Cantonese data did not exhibit a good fit with this model. Similarly, the results of the exploratory factor analysis suggested that sequential and simultaneous factors could apply when describing the factor structure of the English and Punjabi data, but not for the Cantonese data. Implications of these findings are discussed.

Sign in / Sign up

Export Citation Format

Share Document