Good Statistical Fit Research Articles

A polynomial regression model based educational software tool to interpret the internal combustion engine characteristics

Technical education requires regular upgrades in pedagogical methodologies to keep up student’s skill on par with ever demanding job market. This paves the way for creating newer e-learning concepts for classroom to replace or supplement established teaching protocols. In line with this motive, this study deals with the development of an educational software tool to understand the traits of an internal combustion engine. The core of this software tool consists of polynomial regression equations, which in turn was arrived from statistical models using real world experimental data. A MATLAB-based GUI allows the operator to effortlessly interact with the software tool. Upon installation, the software requires the user to define input variables for it to automatically compute data and represent the output data in both visual and tabulated form. The advantage of three-dimensional surface plots for visual representation allows for understating the interactive effect of multiple input parameters on any given output parameter. Overall, average relative error for the model is less than 6%, thus exhibiting a good statistical fit.

Development and validation of a new asthma questionnaire to help achieve a high level of control in school-age children and adolescents

BackgroundMaintaining good asthma control minimizes the risk of exacerbations and lung function decline and is a primary goal of asthma management. The Japanese Pediatric Asthma Guidelines (JPGL) employs different classification criteria for control status from other guidelines, stressing a higher level of control. Based on JPGL, we previously developed a caregiver-completed questionnaire for assessing and achieving best asthma control in preschoolers. In this study, we aimed to develop a questionnaire for school-age children and adolescents. MethodsA working questionnaire comprising 14 items for patients and 34 items for caregivers was administered to 362 asthma patients aged 6–15 years and their caregivers. Separately, physicians filled out a questionnaire to determine JPGL-defined control. Logistic regression analysis was performed to construct a model to predict control levels using data from a randomly selected set of completed questionnaires from two-thirds of the subjects. Validation was performed using the remaining questionnaires. ResultsA set of 7 questions, encompassing self-assessed control status at the time of the visit and in the past month, and nocturnal/early morning asthma symptoms for patients and frequency of asthma symptoms, dyspnea, rescue beta-agonist use, and asthma hospitalization for caregivers, were selected and the 7-item model showed a good statistical fit with AIC of 110.5. The model has been named the Best Asthma Control Test for School Children and Adolescents (Best ACT-S). Best ACT-S scores differed significantly in the hypothetical direction among the groups of different JPGL-defined control levels, step-up/down treatment decisions, and presence/non-presence of exacerbations in the previous year. ConclusionsThe Best ACT-S is a valid questionnaire for children/adolescents aiming for best asthma control.

Statistical clustering of biomass to predict biogas yields

A priori prediction of biogas yields from the biomass composition is difficult owing to the inter and intra variation in the biomass composition. A statistical clustering based on the composition of the biomass has been attempted in this study. Twenty-five biomass feedstocks were clustered into three groups; group 1, rich in extractives; group 2, having a high concentration of holocellulose and group 3, predominantly lignin-rich feedstock. Based on the clustering, two multiple linear regression equations were obtained. The correlations had an R2 > 0.95, showing a good statistical fit. Concentrations of lignin and extractives (L/E) play an important role in the overall biogas yields. Segregation of biomass based on L/E can be used to differentiate feedstocks for efficient biogas production.

Factors affecting shift work sleep disorder in female shift work nurses: Integrating shift workers' coping with stress and transactional stress coping theory.

Nurses have to work different and often irregular shifts to provide care. This puts nurses at health risks, such as sleep impairment. The aim of this study was to test the comprehensive conceptual framework for predicting shift work sleep disorder of female nurses, we used a structural equation model analysis based on shift worker's coping and transactional stress coping theory. This study used a cross-sectional design. We collected the data from three public hospitals and three private hospitals in South Kalimantan, Indonesia, with a total sample of 201 female shift work nurses. Data were collected from February to April 2020. We were also granted permission by the director and the head nurse of these hospitals. After obtaining the informed consent forms, we distributed the online self-report questionnaire using Google Forms. Demographic data were evaluated using descriptive statistical analysis. We used a structural equation model analysis to test the comprehensive conceptual framework for predicting shift work sleep disorder of female shift work nurses. The model's effectiveness in predicting factors contributing to shift work sleep disorder was confirmed by the good statistical fit indicated by the comparative fit index, root-mean-square error of approximation, incremental fit index, and Tucker Lewis Index. This study provides evidences that workload and interpersonal conflict contribute to occupational stress. Workload, interpersonal conflict, and the biological sleep clock influences shift work sleep disorder through mediators of coping strategies and stress.

Structural validity of the Brazilian version of the Sense of Coherence scale (SOC-13) in oral health research: exploratory and confirmatory factor analysis

BackgroundThe Sense of Coherence (SOC) construct has been used worldwide in oral health research, but rigorous factor analyses of the scale are scarce. We aim to test the dimensional structure of the Brazilian short version of the SOC scale with 13 items.MethodsThis study is a secondary analysis of four independent cross-sectional Brazilian studies on oral health, using the 13-items SOC scale. Sample 1 was conducted on 1760 mothers and 1771 adolescents. Sample 2 comprised 1100 adults. Sample 3 had 720 adults and older individuals. Sample 4 comprised 664 adolescent students. Confirmatory Factor Analysis (CFA) was conducted on sample 1 to compare two models: 3-factor versus 1-factor. Because they were refuted, Exploratory Factor Analysis was implemented in samples 2 and 3. Modified models were tested in sample 4 using CFA. All analyses were conducted with MPlus version 7.11.ResultsCFA of sample 1 resulted in an unacceptable fit (RMSEA = 0.12;CFI = 0.78; TLI = 0.73; and WRMR = 3.28) for 1-factor model and 3-factor (RMSEA = 0.10; CFI = 0.87; TLI = 0.84; and WRMR = 2.50). The EFA on samples 2 and 3 showed, respectively, two eigenvalues greater than 1 (4.11 and 1.56) and (4.32 and 1.42), but the scale items soc1, soc2 and soc3 formed an uninterpretable second factor. Another CFA, using sample 4, showed acceptable model fit after removing those three items and also soc11 (RMSEA = 0.05; CFI = 0.98; TLI = 0.99; and WRMR = 0.71).ConclusionThe results indicate that the SOC-13 scale needs further adjustments. The one-factor model with nine items showed a good statistical fit, but the implications of excluding items should be further investigated, considering the scale's content validity, cross-cultural adaptation and theoretical background.

Effect of different interventions for latent tuberculosis infections in China: a model-based study

BackgroundTuberculosis (TB) has a serious impact on people’s health. China is one of 30 countries that has a high TB burden. As the currently decreasing speed of the incidence of TB, the WHO’s goal of “End TB Strategy” is hard to achieve by 2035. As a result, a SEIR model that determines the impact of different tuberculosis preventive treatments (TPTs) in different age groups, and the effect of different interventions on latent TB infections (LTBIs) in China is developed.MethodsA Susceptible-Exposed-Infectious-Recovered (SEIR) model was established. Goodness-of-fit tests were used to assess model performance. Predictive analysis was used to assess the effect of different interventions on LTBIs and achieving the goals of the “End TB Strategy”.ResultsThe Chi-square test indicated the model provided a good statistical fit to previous data on the incidence of TB (χ2 = 0.3085, p > 0.999). The 1HP treatment regimen (daily rifapentine + isoniazid for 4 weeks) was most effective in reducing the number of TB cases by 2035. The model indicated that several strategies could achieve the 2035 target of the “End TB Strategy”: completion of active case finding (ACF) for LTBI and TPT nation-wide within 5 years; completion of ACF for LTBIs and TPT within 2 years in high-incidence areas; completion of TPT in the elderly within 2 years; or introduction of a new vaccine in which the product of annual doses and vaccine efficiency in the three age groups above 14 years old reached 10.5 million.ConclusionThe incidence of TB in China declined gradually from 2005 to 2019. Implementation of ACF for LTBIs and TPT nation-wide or in areas with high incidence, in the elderly, or administration of a new and effective vaccine could greatly reduce the number of TB cases and achieve the 2035 target of the “End TB Strategy” in China.

Applying reactance theory to study consumer responses to COVID restrictions: a note on model specification

PurposeThe purpose of the study is to highlight the relevance of reactance theory for modeling consumer responses to COVID restrictions. The study also aims to critically evaluate the appropriateness of the most established reactance model (the intertwined model) for studying reactance specifically in relation to freedom threats arising from measures aimed at combatting the spread of the pandemic.Design/methodology/approachFollowing a conceptual analysis of the intertwined model of reactance, structural equation modeling is applied to Rain's (2013) meta-analytic data to compare the model to alternative model specifications.FindingsThe analysis reveals both conceptual and statistical shortcomings of the intertwined model of reactance in its current/traditional form. It also draws attention to other model specifications that provide just as good statistical fit and offer promising alternative ways of modeling reactance in a COVID context.Originality/valueThe study is the first attempt to explicitly discuss conceptual and statistical problems associated with the most widely accepted model of reactance, illustrate these issues with specific reference to consumer reactions to COVID restrictions, identify alternative promising model specifications and suggest a respecification of the intertwined model.

Adsorption Isotherms of Dimethyl Phthalate (DMP), DiEthyl Phthalate (DEP), Dibutyl Phthlate (DBP) and Benzyl Butyl Phthalate (BBP) Compounds in Synthetic and Sembrong River Sediment

Concentrations of Dimethyl Phthalate (DMP), Diethyl Phthalate (DEP), Dibutyl Phthlate (DBP), and Benzyl Butyl Phthalate (BBP) analysis proved present in Sembrong river sediment. These findings occurred during a certain extent of time depending on the properties of sediment and phthalates esters compound. Therefore, adsorption isotherm conducted using optimization method using manipulating variables of initial concentrations and contact times. The results analysis showed higher percentage adsorption recorded in DMP, DEP, DBP and BBP compounds by synthetic sediment which is 88.3%,89.6%,89.0% and 94.8% whereas river sediment is 87.4%,88.4%, 87.1% and 93.4%. The contributing factors aided in the adsorption of PAEs onto synthetic sediment are higher organic content (OC), smaller particle size having larger surface area and higher molecular weight of PAEs. Competitive adsorption occurred between DMP, DEP and DBP as they are low molecular weight compounds. Adsorption isotherms models by Langmuir and Freundlich isotherms models showed separation factor, RL values are favourable from 0.11 to 0.69. Good statistical fit R2 ranged from 0.97 to 0.99 of a good fit. Therefore, in conclusion, Sembrong river sediment is an adsorbent sediment storage of PAEs compounds.

Psychometric Validation of the Korean Version of the Heart Failure-Specific Health Literacy Scale.

The goals of heart failure (HF) management include lowering adverse outcomes through optimal self-care. Health literacy has a significant role in ensuring optimal and effective HF self-care. Unfortunately, the proportion of low health literacy in HF patients is higher than those with other chronic diseases as well as the general population. Thus, healthcare providers need to screen HF patients. The HF-Specific Health Literacy Scale is a specific questionnaire to screen the health literacy in HF patients; however, the Korean version has not yet been validated. The study aimed to investigate the reliability and validity of the Korean version of the HF-Specific Health Literacy Scale. We adopted a cross-sectional study design with 386 outpatients with chronic HF from a tertiary care hospital in Cheonan, South Korea. Data were collected from April to December 2020. The exploratory factor analysis revealed a three-factor (functional, communicative, and critical health literacy) 12-item structure of the scale. The result of the confirmatory factor analysis confirmed a good statistical fit for the data. The Korean version of the tool demonstrated satisfactory convergent and discriminant validity. The criterion validity analysis revealed significant correlation with general health literacy and knowledge about HF. The Korean version of the tool for 12 items had adequate overall internal consistency. The Korean version of the HF-Specific tool can be easily utilized for assessing health literacy level for HF patient education. Moreover, the tool can help healthcare providers develop strategies for promoting HF patients' health literacy.

Comparison of three indices of relative income deprivation in predicting health status

Relative income deprivation (RID) is a known risk factor for poor health. Previous research has proposed several measures to assess RID, e.g., Income Rank and the Yitzhaki Index. Hounkpatin et al. (2020) presented a new approach – the CR˜i index – to account for the observation that individuals are more sensitive to the differences in incomes of others who are closer to them, rather than to comparisons with incomes of others far above them. Using a Japanese nationwide cohort of older adults (n = 62,438; mean [SD] age: 73.0 [5.6] years), this study compared the performance of alternative indices of RID in predicting health outcomes (depressive symptoms, functional capacity, and self-rated health), as well as the use of alternative CR˜i index weights (α weight range: −0.9 to 0.9). When 0<α<1, higher income differences lead to a more significant increase in relative deprivation, while when −1<α <0, excessively high incomes contribute less to the relative deprivation of lower income individuals in the same reference group. Results showed that all measures of relative income deprivation were associated with deteriorating mental and physical health among older Japanese adults. However, while the CR˜i index consistently outperformed the Yitzhaki Index, this did not hold true invariably when compared to the Income Rank – depending on the health outcome and the reference group. Also, while negative α parameters showed a good statistical fit in most models, the findings were not conclusive – the best-fitting CR˜i weight parameters ranged from −0.9 to 0.9. Therefore, a clear direction for the contribution of higher incomes to relative deprivation could not be established based on the study population.

Assessing Well-Being Among Aging Women: Observations From the Women's Health Initiative

Our aim was to develop a profile of well-being preserving the ability to estimate differential effects of both hedonic and eudaemonic dimensions of well-being on health outcomes. Numerous indicators of well-being from over 80,000 aging women included hedonic (enjoyment of life, happiness, satisfaction with life, quality of life) and eudaemonic (personal growth, purpose in life, environmental mastery, control, self mastery) dimensions. Using latent class analysis, we identified groups of women with distinct profiles of well-being. A four-class solution had both good statistical fit and made conceptual sense. Class 1 (n=9,146, 11%) had the lowest scores on hedonic and eudaemonic indicators, while Class 4 (n=38,246 47%) had the highest levels of all well-being indicators. Class 2 (n=7,106, 9%) had higher hedonic and lower eudaemonic scores and Class 3 (n=26,650, 33%) had lower hedonic and higher eudaemonic scores. These classes form a well-being profile useful for estimating differential effects on health outcomes.

Testing a Middle -Range Theory of Heart Failure Self-Care.

A middle-range theory of heart failure self-care, derived from the self-care deficit theory of nursing, was tested among 175 Arab American older adults with heart failure. The middle-range theory achieved good statistical fit, but not all hypothesized relationships were supported. Specifically, conceptualizing basic conditioning factors as a single latent variable was not supported. However, individual factors of depression, social support, and time living with heart failure had a direct effect on both self-care agency and quality of life. Understanding predictors and outcomes of self-care within a theoretical framework is essential in caring for patients with heart failure.

Deterioration of Health-Related Quality of Life After Withdrawal of Risankizumab Treatment in Patients with Moderate-to-Severe Plaque Psoriasis: A Machine Learning Predictive Model.

IntroductionRisankizumab has demonstrated efficacy in treating moderate-to-severe psoriasis. The phase-3 IMMhance trial (NCT02672852) examined the effect of continuing versus withdrawing from risankizumab treatment on psoriasis severity, including the Psoriasis Area and Severity Index (PASI) and static Physician Global Assessment (sPGA). However, the effect of withdrawal on health-related quality of life (HRQL) was not assessed. Therefore, this study was conducted to evaluate the impact of risankizumab withdrawal on HRQL measured by the Dermatology Life Quality Index (DLQI). Because DLQI was not measured beyond week 16 in IMMhance, a machine learning predictive model for DLQI was developed.MethodsA machine learning model for DLQI was fitted using repeated measures data from three phase-3 trials (NCT02684370, NCT02684357, NCT02694523) (pooled N = 1602). An elastic-net algorithm performed automated variable selection among candidate predictors including concurrent PASI and sPGA, demographics, and interaction terms. The machine learning model was used to predict DLQI at weeks 28–104 of IMMhance among patients re-randomized to continue (N = 111) or withdraw from (N = 225) risankizumab after achieving response (sPGA = 0/1) at week 28.ResultsThe machine learning predictive model demonstrated good statistical fit during tenfold cross-validation and external validation against observed DLQI at weeks 0–16 of IMMhance (N = 507). Predicted improvements in DLQI from baseline were lower in the withdrawal versus the continuation cohort (mean DLQI change at week 104, −5.9 versus −11.5, difference [95% CI] = 5.6 [4.1, 7.3]). Predicted DLQI deteriorated more extensively than PASI (49.7% versus 36.4%) after treatment withdrawal.ConclusionsThe predicted DLQI score deteriorated more rapidly after risankizumab withdrawal than the PASI score, an objective measure of disease. These findings suggest that the deterioration in HRQL reflects more substantial impacts after risankizumab discontinuation than those measured by PASI only.Supplementary InformationThe online version contains supplementary material available at 10.1007/s13555-021-00550-8.

OPTIMAL LIQUIDATION TRAJECTORIES FOR THE ALMGREN–CHRISS MODEL

We consider an optimal liquidation problem with infinite horizon in the Almgren–Chriss framework, where the unaffected asset price follows a Lévy process. The temporary price impact is described by a general function that satisfies some reasonable conditions. We consider a market agent with constant absolute risk aversion, who wants to maximize the expected utility of the cash received from the sale of the agent’s assets, and show that this problem can be reduced to a deterministic optimization problem that we are able to solve explicitly. In order to compare our results with exponential Lévy models, which provide a very good statistical fit with observed asset price data for short time horizons, we derive the (linear) Lévy process approximation of such models. In particular we derive expressions for the Lévy process approximation of the exponential variance–gamma Lévy process, and study properties of the corresponding optimal liquidation strategy. We then provide a comparison of the liquidation trajectories for reasonable parameters between the Lévy process model and the classical Almgren–Chriss model. In particular, we obtain an explicit expression for the connection between the temporary impact function for the Lévy model and the temporary impact function for the Brownian motion model (the classical Almgren–Chriss model), for which the optimal liquidation trajectories for the two models coincide.

A Nonparametric Bayesian Methodology for Synthesizing Residential Solar Generation and Demand Data

The uptake of behind-the-meter distributed energy resources in low-voltage distribution networks has progressed to a point where phenomena including overvoltage and reverse power flow are emerging. The limited nature of supervision and control on these networks require novel tools for operation and planning in this environment. In this paper, we propose a methodology for synthesizing stochastic demand and generation profiles for unobserved customers with rooftop PV, called prosumers . The proposed model bridges the gap between the limited available empirical data, and the large amount of high-quality, stochastic demand and generation data required for probabilistic analysis. The approach employs clustering analysis and a Dirichlet-categorical hierarchical model of the features of unobserved prosumers. Based on the data of clusters of prosumers, Markov chain models of demand and generation profiles are constructed from empirical data, and synthetic demand profiles are subsequently sampled from these. The sampled traces are cross-validated and show a good statistical fit to the observed data. Two case studies are considered to confirm the validity of the proposed methodology. The first studies the impact of behavioral differences on the synthetic demand profiles, while the second looks at the impact of varying solar generation penetration on demand profiles.

Applying the Concept of Successful Aging to Thailand

Objectives: This research aims to examine the components of Rowe and Kahn’s model of successful aging as applied to the Thai population who are categorized as: young-old (60-69), old-old (70-79), and oldest-old (80 and over). Method: The study is based on the Survey of Older Persons in Thailand of the National Statistical Office, 2011 (n = 24,433). Confirmatory Factor Analysis is used to test whether Rowe and Kahn’s model, which comprises three domains - avoiding disease and disability, maintaining physical and cognitive functioning, and continuing engagement with life, is confirmed by these data. Multiple group analysis is performed to determine factor loading differences among age groups. Results: Rowe and Kahn’s model has a good statistical fit when applied to data of the current study. Factor loadings for three age groups is statistically different (p&lt;0.001). Of the three domains, continuing engagement with life is the domain most strongly associated with successful aging, particularly among the old-old and the oldest-old. Implications: Programs to support health and social engagement should be promoted and enhancing active engagement should be a priority since older Thai people have low levels of social participation.

Developing and Validating the Functional, Communicative, and Critical Health Literacy Questionnaire among the Iranian General Population.

ObjectivesMany scales have been developed to assess health literacy. These scales have been developed extensively in clinical populations in developed countries. Our study describes the process of development and validation of the Health Literacy Questionnaire to measure health literacy in the general Iranian population.MethodsWe followed the scale development process outlined by Schwab in 1980. After a comprehensive review of the research published on Nutbeam’s definition of health literacy and its measurement, two focus groups were considered to generate a pool of items. We then assessed the face validity and content validity of the items. The final version of the questionnaire was subject to independent exploratory and confirmatory factor analysis. The internal consistency of the questionnaire was evaluated using Cronbach’s alpha. Finally, construct validity was assessed using bivariate analysis between the total scores on the scale and theoretically relevant variables.ResultsThe results of exploratory factor analysis revealed a three-factor 14-item structure for the scale. Confirmatory factor analysis provided a good statistical and conceptual fit for the data. The analysis of the internal consistency of the scale was satisfactory (α = 0.798). Further, health literacy was significantly associated with participants’ age, educational level, self-rated household income, smoking history, self-rated health, and self-rated health literacy.ConclusionsThe results we obtained indicate that this newly constructed health literacy tool is highly valid and reliable. Prospective studies are required to evaluate the predictive validity of the scale with regard to health outcomes.

Spatial and temporal variation of the vegetation of the semiarid Puna in a pastoral system in the Pozuelos Biosphere Reserve.

This study aimed to analyze the spatial and temporal variation of the vegetation in the northern Argentine Puna, utilizing both field sampling and remote-sensing tools. The study was performed within the Pozuelos Biosphere Reserve (Jujuy province, Argentina), which aims to generate socio-economic development compatible with biodiversity conservation. Our study was designed to analyze the dynamics of the Puna vegetation at local scale and assess and monitor the seasonal (dry and wet seasons), interannual, and spatial variation of the vegetation cover, biomass, dominant species, and vegetation indices. Ten vegetation units (with differences in composition, cover, and high and low stratum biomass) were identified at our study site. The diversity of these vegetation units correlated with geomorphology and soil type. In the dry season, the vegetation unit with greatest vegetation cover and biomass was the Festuca chrysophylla grassland, whereas in the wet season, the units with greatest cover and biomass were vegas (peatlands) and short grasslands. The Festuca chrysophylla grasslands and short grasslands were located in areas with clay soils, except peatlands, associated with valleys and coarse-texture soils. The vegetation indices used (NDVI, SAVI, and MSAVI2) were able to differentiate functional types of vegetation and showed a good statistical fit with cover values. Our results suggest that the integrated utilization of remote-sensing tools and field surveys improves the assessment of the Puna vegetation and would allow a periodic monitoring at production unit scale taking into account its spatial and temporal variation.

Meteor Radar Estimations of Gravity Wave Momentum Fluxes: Evaluation Using Simulations and Observations Over Three Tropical Locations

AbstractMeteor radars are widely used to study gravity wave (GW) variances and their momentum fluxes at the altitudes where meteor counts are sufficient to yield good statistical fits to the data. These radars provide hourly zonal and meridional wind observations round the clock in 80‐ to 100‐km height domain. However, the capability of meteor radars in estimating GW momentum fluxes should be evaluated before they can be used for any research applications. In this regard, the present study evaluates the meteor radar observations of GW momentum fluxes obtained from Thumba (8.5°N, 77°E; 2006–2015), Kototabang (0.2°S, 100.3°E; 2002–2017), and Tirupati (13.63°N, 79.4°E; 2013–2018) using three‐dimensional wind field simulations, which include specified tidal, planetary and GW fields. A modified composite day analysis is adopted to estimate the GW momentum fluxes, which also accounts for the tidal and planetary wave contributions. The results showed that the retrieved and specified GW momentum fluxes agree very well over Tirupati followed by Thumba and Kototabang. It is noted that the agreement between the retrieved and simulated fields depend on the number of meteor detections used in the analysis. After evaluating the meteor radar retrieved GW momentum fluxes by employing simulations, their interannual variability and climatologies over the three observational locations are constructed. The significance of the present study lies in evaluating the capability of three meteor radars located in equatorial and low‐latitudes in estimating GW momentum fluxes by employing three‐dimensional wind field simulations with specified mean winds, tidal, planetary, and GW fields.

Assessing the Suitability of The Adapted Scientific Epistemic Beliefs Questionnaire in Namibia

This study adapted and validated a questionnaire for assessing Namibian students’ scientific epistemic beliefs and explore their relation to achievement in science, gender, grade and socioeconomic status. The study was correlational, with a sample of 944 (45% male; 55% female) Grades 11 and 12 students with the mean age M=17.9, SD= 1.4. A Likert scale questionnaire comprised of four dimensions of beliefs: source; certainty; development; and justification of scientific knowledge. The adapted questionnaire had good reliability with the Cronbach’s alpha of subscales ranging from .80 to .83 and the overall reliability of .70. Model fit analysis yielded good statistical fit with Chi-square ratio to degrees of freedom= 2.71, RMSEA= 0.043, CFI= 0.95, TLI= 0.94, SRMR= .032. The overall regression model was significant F (4, 939) =8.218, p &lt;.001, R2= .034. Two dimensions, certainty (b=.154, p&lt;.001) and justification (b=.100, p&lt;.05) statistically significantly predicted achievement in science. There was statistically significant difference in beliefs about source in terms of gender and grade and about certainty in terms of grade. The results show that the questionnaire works well with the Namibian sample used given the good model fit for the data and reliability. Results are discussed in terms of 21st century skills development.