Research Design and Statistical Analysis Using SPSS Training Course

Participants can define variables and measurement rules suitable for statistical analysis.

• Constructs operationalised through observable indicators, questions and administrative measures
• Nominal, ordinal, interval and ratio measurement levels classified correctly
• Dependent, independent, control, mediator and moderator variables distinguished
• Question wording assessed for ambiguity, double-barrelled items and response bias
• Likert scales designed with balanced response categories and anchored labels
• Codebooks specify variable names, labels, values, formats and missing codes

Participants can select and justify sampling methods for a defined study population.

• Target populations, sampling frames and eligibility criteria documented explicitly
• Simple random, systematic and stratified sampling methods compared
• Cluster and multistage samples assessed for cost and precision
• Convenience, purposive and quota sampling limitations identified clearly
• Sample size linked to precision, power, effect size and resources
• Nonresponse patterns examined for likely coverage and selection bias

Participants can specify data collection controls that protect quality, privacy and consent.

• Consent statements distinguish participation, data use, storage and withdrawal
• Data minimisation principles applied when designing collection instruments
• Pilot testing identifies unclear questions, timing issues and routing errors
• Interviewer protocols reduce inconsistent prompting and recording practices
• Validation rules prevent impossible dates, ranges and incompatible responses
• Data collection logs record versions, exceptions, corrections and fieldwork events

Participants can create a structured SPSS data file from a codebook.

• IBM SPSS Statistics windows, viewers and output management configured
• Variable View used to define names, labels, types and formats
• Value labels and user-missing values assigned consistently
• Data View records cases, variables and imported observations correctly
• Measurement levels set to support appropriate SPSS procedures
• Portable file formats compared: SAV, CSV, XLSX and syntax files

Participants can import and audit datasets before conducting analysis.

• CSV and Excel imports checked for headers, types and encoding
• Frequencies reveal invalid codes, duplicates and unexpected category values
• Descriptive statistics identify impossible ranges and suspicious extreme values
• Case identifiers inspected before correcting or removing duplicate records
• Missing values distinguished from valid zeroes, blanks and refusals
• Data cleaning decisions recorded in an auditable processing log

Participants can transform variables and record reproducible SPSS processing steps.

• COMPUTE commands create scores, rates, ratios and derived variables
• RECODE procedures group categories while retaining documented coding rules
• IF conditions apply transformations only to eligible cases
• DATE functions calculate durations, periods and age variables
• DO IF structures manage conditional multi-step data transformations
• Syntax files document repeatable imports, cleaning and analysis procedures

Participants can combine datasets safely and prepare analysis-ready files.

• ADD FILES combines records with matching variable definitions
• MATCH FILES merges case-level records using unique identifiers
• One-to-one and one-to-many merge risks identified before processing
• Variable naming conventions preserve source, wave and questionnaire meaning
• Analysis datasets separated from source files and archived securely
• Data dictionaries updated after merges, recodes and calculated fields

Participants can summarise distributions and select charts that communicate patterns accurately.

• Frequencies, percentages and valid percentages reported with appropriate denominators
• Mean, median and mode compared for different distribution shapes
• Range, variance, standard deviation and interquartile range interpreted
• Histograms and boxplots reveal skewness, outliers and distribution gaps
• Bar charts, pie charts and tables selected for categorical variables
• Chart titles, axes and data labels edited for decision audiences

Participants can interpret statistical inference without overstating the evidence.

• Sampling distributions explain uncertainty around sample-based estimates
• Confidence intervals communicate plausible population values and estimate precision
• P-values interpreted against pre-specified significance thresholds carefully
• Type I and Type II errors linked to decision consequences
• Statistical power relates sample size, effect size and error rates
• Practical significance distinguished from statistical significance in reporting

Participants can select and interpret tests for differences between two groups.

• Independent-samples t-tests compare means for separate respondent groups
• Paired-samples t-tests assess change across matched observations
• Normality and variance assumptions checked before parametric comparisons
• Levene's test informs interpretation of equal-variance test results
• Cohen's d quantifies standardised differences between group means
• Mann-Whitney U and Wilcoxon tests address ordinal or non-normal data

Participants can analyse mean differences across multiple groups and report follow-up tests.

• One-way ANOVA compares mean outcomes across three or more groups
• Factorial ANOVA tests main effects and interaction effects
• Homogeneity of variance assessed using Levene's test
• Post-hoc comparisons identify groups responsible for overall differences
• Eta squared reports variance explained by group membership
• Kruskal-Wallis tests compare independent ordinal or non-normal group distributions

Participants can test associations between categorical variables using contingency tables.

• Crosstabs display joint distributions, row percentages and column percentages
• Pearson chi-square tests assess association between categorical variables
• Expected cell counts checked before accepting chi-square results
• Fisher's exact test handles small contingency table samples
• Phi and Cramer's V describe association strength appropriately
• Standardised residuals identify cells contributing most to association results

Participants can quantify and interpret associations between two measured variables.

• Scatterplots assess direction, form, clustering and influential observations
• Pearson correlation measures linear association between continuous variables
• Spearman rank correlation suits ordinal or non-normal variables
• Correlation coefficients interpreted with confidence intervals and sample size
• Correlation does not establish causation without design-based supporting evidence
• Restriction of range and outliers can distort correlation estimates

Participants can build and interpret linear regression models for continuous outcomes.

• Regression equations estimate continuous outcomes from multiple predictor variables
• Unstandardised and standardised coefficients answer different interpretation questions
• R squared and adjusted R squared assess explained outcome variation
• Residual plots assess linearity, homoscedasticity and influential cases
• Variance inflation factors identify problematic multicollinearity among predictors
• Hierarchical regression tests incremental contribution of predictor blocks

Participants can model binary outcomes and explain odds-based results to decision-makers.

• Binary logistic regression models yes-no and event outcomes
• Odds ratios translate predictor effects into relative likelihood changes
• Reference categories determine interpretation of categorical predictor coefficients
• Classification tables compare predicted outcomes with observed outcomes
• Hosmer-Lemeshow results considered alongside substantive model diagnostics
• Predicted probabilities support risk segmentation and targeted interventions

Participants can evaluate multi-item measures and identify underlying dimensions.

• Cronbach's alpha assesses internal consistency of multi-item scales
• Item-total correlations identify items weakening scale reliability
• Reverse-coded items checked before calculating composite scale scores
• Kaiser-Meyer-Olkin measure assesses factor analysis sampling adequacy
• Principal axis factoring identifies latent dimensions among correlated items
• Rotated factor loadings support interpretable scale and construct decisions

Participants can diagnose data issues and choose proportionate treatment methods.

• Missing completely at random distinguished from systematic missingness
• Listwise and pairwise deletion compared for analytic consequences
• Multiple imputation principles introduced for plausible missing-data treatment
• Standardised scores and boxplots identify univariate outliers
• Mahalanobis distance screens multivariate outliers in regression datasets
• Sensitivity analysis documents how treatment choices affect key conclusions

Participants can present analysis findings with appropriate statistical and managerial interpretation.

• Results sections separate findings, interpretation, limitations and recommendations
• APA-style statistical notation reports tests, degrees of freedom and p-values
• Tables present denominators, estimates, confidence intervals and effect sizes
• Executive summaries state the decision question before analytical detail
• Limitations address sampling, measurement, missing data and causal claims
• Recommendations connect evidence strength to feasible management actions

Participants can build, review and present a complete research design, SPSS analysis and evidence-based findings package.

• Research question, conceptual model and hypotheses finalised for selected case
• Codebook and SPSS dataset reviewed for measurement and coding quality
• Cleaning syntax, transformations and analysis decisions documented transparently
• Descriptive, inferential and model outputs selected against research questions
• Findings report includes assumptions, limitations, effect sizes and recommendations
• Peer review tests methodological logic, interpretation and presentation clarity