How to Choose Research Questions and Tests for the HBC203 ECA
9 min read
You have an 872-row dataset, a 2,000-word limit and a brief that wants three research questions, each answered with a different statistical test. The question that costs the most marks is not the jamovi click-path. It is whether your HBC203 ECA research questions can be answered at all with the columns you were given.
Quick answer: Pick each research question by the type of variables it links. Two group means call for a t-test, three or more call for a one-way ANOVA, two continuous scales call for correlation and simple regression, and two categorical variables call for chi-square. Keep every question inside one block of the dataset, and use a different test for each.
What markers actually check in the HBC203 ECA
In the July 2026 semester brief, the report is worth 100 marks, all parts compulsory, with a 2,000-word target for the text. The dataset mixes two survey years of a quiet-quitting study, and each year holds different columns. That design is deliberate. It tests whether you notice which variables sit in which block.
Markers read three things first: whether each question is answerable, whether the named test fits the variable types, and whether the screenshots are legible. Style comes later. A beautifully written question about a variable that does not exist in that block scores lower than a plain question that works.
Step 1: List your variables by type before you write a question
Do this on paper before opening the file. Write each column, which block it belongs to, and whether it is a rating, a 0 to 100 score, a ranking or a category. Most students skip this and invent questions from the scenario text, then discover in jamovi that the columns do not line up.
- Note which columns exist only in the earlier survey year and which only in the later one.
- Mark each as nominal, ordinal or continuous. A one-to-five rating is ordinal in theory, so decide whether you will treat it as continuous and be ready to justify it.
- Circle any column that appears in both blocks. These are your only candidates for a comparison across years.
- Check whether respondents can be matched between years. In an anonymous survey they usually cannot.
Decision table: question type, test, jamovi menu and APA line
This table covers the methods HBC203 teaches. Methods outside the course, such as multiple regression, do not earn marks here, so leave them out even if you know them.
| Question type | Test | jamovi menu | APA line pattern |
|---|---|---|---|
| Do two separate groups differ on a score? | Independent-samples t-test with Cohen’s d | T-Tests, Independent Samples T-Test | t(df) = value, p = value, d = value |
| Does the same group differ across two conditions? | Dependent-samples t-test | T-Tests, Paired Samples T-Test | t(df) = value, p = value, d = value |
| Do three or more groups differ on a score? | One-way between-subjects ANOVA, Tukey, eta squared | ANOVA, One-Way ANOVA | F(df1, df2) = value, p = value, η² = value |
| Does one score change across three or more time points for the same people? | Repeated-measures ANOVA | ANOVA, Repeated Measures ANOVA | F(df1, df2) = value, p = value |
| Are two continuous scales related? | Pearson correlation, then simple regression | Regression, Correlation Matrix; Linear Regression | r(df) = value, p = value; b = value, R² = value |
| Is one category split as expected? | Chi-square goodness of fit | Frequencies, N Outcomes | χ²(df, N = value) = value, p = value |
| Are two categorical variables associated? | Chi-square test of independence | Frequencies, Contingency Tables, Independent Samples | χ²(df, N = value) = value, p = value |
Step 2: Write three questions that can fail honestly
A good question names two variables, a population and a direction you can predict from the literature. “Is intention to stay lower in the later survey year than in the earlier one?” works. “Why do employees quietly quit?” does not, because no test in the course answers a why.
Match each question to a different method
The brief asks for different methods across the three questions, so plan the set together. One comparison of two groups, one relationship between two scales, and one test involving a categorical variable gives a clean spread. Our HBC203 ECA model extract shows one such set with clearly labelled illustrative numbers, for reference only.
Keep each question inside one block
If the 2024 block has a happiness category and an intention score, a question about pride belongs to the other block. Mixing them forces you to compare people who never answered the same item. Subset the data first, then write the question to fit what remains.
Step 3: Justify recoding and subsetting in plain sentences
The brief asks you to defend any recoding or filtering. One sentence per decision is enough: what you changed, why, and what it cost. Examples include stacking two intention columns into one with a year label, filtering to a single year, or collapsing a ranking into “ranked first” versus “not ranked first” so a chi-square test becomes possible.
Collapsing loses information, and a marker will expect you to say so. A recode you cannot justify is a recode you should not make.
Screenshot legibility: the quiet mark-loser
A screenshot that cannot be read is treated as missing evidence. Most students paste the entire jamovi window at a size where the numbers blur once the Word file is shrunk to a page.
| Common problem | What the marker sees | Fix |
|---|---|---|
| Whole window pasted | Tiny numbers, hard to match to the text | Crop to the results table only |
| Test name cropped out | No proof of which test was run | Include the table title and variable names |
| Output on a second page from its question | Unclear which RQ it supports | Place each image directly under its RQ heading |
| Missing effect size | Incomplete APA line | Tick Cohen’s d or eta squared before taking the screenshot |
Zoom the jamovi output to at least 125% before capturing, and read the pasted image at 100% in your document before you submit.
Word-count traps
The 2,000 words count text only. References, appendices and jamovi output do not count, but a long table you typed yourself does. Excess words are simply not graded, which means a bloated introduction can push your conclusion past the cut-off.
- Spend about 250 words on the introduction, about 400 on each research question and about 200 on the conclusion.
- Put the full jamovi output in the screenshots, not in prose.
- Report each statistic once. Do not repeat it in the discussion.
- State your word count honestly, using your own word processor’s count.
Checklist before you submit
- Each research question names two variables from one block of the data.
- Each question uses a different course method.
- Each has a short literature review with at least two academic sources in APA 7.
- Each has a prediction that comes from those sources, not from the data.
- Test names are justified by variable type, not by habit.
- Screenshots are cropped, legible and placed under their question.
- APA lines report the test statistic, degrees of freedom, p value and an effect size.
- Discussion states whether the prediction was supported, compares with the literature, gives an implication, and names a limitation with a fix.
- The AI declaration matches SUSS rules and what you actually did.
Common mistakes we see
Three come up every semester. The first is running a paired t-test on two survey years of anonymous respondents. The second is reporting only p, with no effect size, which the brief expects. The third is writing a discussion that restates the results instead of asking whether the literature agrees.
My view: the weakest reports are not the ones with wrong numbers. They are the ones whose questions were never checked against the columns. Ten minutes with the variable list saves hours later.
When a second pair of eyes helps
Students come to us at two points: before writing, to check the three questions fit the data, and after the jamovi run, to check the APA lines and screenshots. Our statistics assignment help service reviews plans and outputs. For the report layout itself, the report writing help page covers structure and referencing. SUSS modules are listed under SUSS assignment help, and a past HBC203 brief appears on this HBC203 ECA question page.
Hui Yan Lee, who holds a Master’s in Industrial-Organizational Psychology and works on employee engagement and retention topics, is the kind of specialist to ask about engagement literature. See her expert profile.
FAQs
How many research questions does the HBC203 ECA need?
The July 2026 brief asks for three research questions, and each must be answerable with the dataset. Each one needs a different HBC203 method, so plan the set together before you write any of them. Always confirm the number against your own Canvas brief.
Can I use a paired t-test to compare 2024 and 2025?
Usually not. The survey is anonymous, so a 2024 respondent cannot be matched to a 2025 respondent. Two independent groups call for an independent-samples t-test. Explain this choice in one sentence, because markers look for it.
Do I need to test assumptions in jamovi?
The brief says assumption testing is not required, so do not spend words on it. You can note one assumption briefly if it supports a decision, but your word budget is better spent on the discussion.
Is multiple regression allowed?
Only methods taught in HBC203 earn marks, and that means simple linear regression, not multiple regression. Adding extra predictors will not help your score, so mention them only as future work.
Which software should I use for the analysis?
The brief requires jamovi, and screenshots of the output are part of the assessment. Using SPSS or Excel for the main analysis would not match the requirement, even if the numbers agree.
What happens if I go over 2,000 words?
Words beyond the target are not graded. That means any content after the cut-off, often your conclusion, earns nothing. Trim the introduction first, and keep your final sections inside the limit.
If your deadline is close and you are unsure your three questions will survive marking, send your variable list and draft questions through our SUSS ECA help page for a plan check. See what other students say on the reviews page. We provide guidance and model structures, and the work you submit must be your own.
The post How to Choose Research Questions and Tests for the HBC203 ECA appeared first on Singapore Assignment Help.
