Results-chapter and executive-summary generator: from free input

Generator

Study details

Hypotheses and findings

One row for each hypothesis or analysis. Enter the numbers as the calculator or SPSS printed them – the generator fixes the leading zero according to APA (section 6.36: p, r, β without a zero; t, F, d with a zero), marks a negative value with a minus sign, and assembles the report sentence. What was not filled in does not appear, and the checklist at the end reminds you.

Conclusions, limitations and recommendations (for the abstract and the executive summary)

Guide

What the generator does – and what it does not

The first generator, from calculator outputs, assembles a results chapter from files that the site's calculators produced. This generator starts from the other side – from a form: someone who analyzed in SPSS, in R or in another calculator, or who also wants an abstract and an executive summary, types in the study details and the numbers for each hypothesis, and gets three documents in fixed wording: a results chapter (following APA, 2020, section 3.7 – the order of the hypotheses, a report sentence for each, and a checklist), an abstract (in APA structure – sections 2.9 and 3.3: objective, participants, method, results, conclusions – in a single paragraph of up to 250 words) and an executive summary for the client (in plain language, following the site convention below). All the sentences are built from fixed templates out of what you typed: the generator does not invent numbers, does not interpret and does not fill in – what was not filled in does not appear, and the checklist says what is missing. The site convention: no artificial intelligence and no automatic wording beyond the template; the result is a first draft that you edit.

What the generator does not do. It does not compute: the numbers come from the calculators (see the table on the reporting guidelines page – what must be in the sentence for each test) or from your software. It does not write a discussion: according to APA (section 3.8) the discussion interprets, compares with the literature and discusses limitations – the generator gives it only the opening the section requires ("a clear statement of the support or nonsupport for all hypotheses") and the limitations you wrote. And it does not replace judgment: "supported" or "not supported" is your decision, which the generator only words.

The abstract according to APA

APA (2020, section 2.9) defines an abstract as "a brief, comprehensive summary of the contents of the paper", usually up to 250 words, in a single paragraph without indentation – or in a structured format with labels (Objective, Method, Results, Conclusions), as the journal requires. Section 3.3 says what a good abstract should be and what it should contain. The qualities: accurate (reflects the paper and does not add to it), nonevaluative ("report rather than evaluate"), coherent and readable (verbs rather than noun equivalents, active voice; past tense for what was measured, present tense for conclusions), and concise – "make each sentence maximally informative, especially the lead sentence", without repeating the title, only four or five important concepts or findings. The content for an empirical article: the problem under investigation in one sentence, including the hypotheses; the participants or data sources with their characteristics; the features of the method – design, analysis strategy, data collection, sample size, measures; the main findings – in a quantitative analysis "effect sizes and confidence intervals in addition to significance levels, wherever possible"; and the conclusions and implications. The generator fills in the five components in this order from the fields in the form, counts words and warns above 250, and also offers the structured version with the labels (in APA – in bold italics; here bold).

Executive summary – the site convention

APA has no genre called an "executive summary"; what it has is the abstract (3.3) and the discussion (3.8). The site convention, following the user's decision: the executive summary is built from the same components as the academic abstract – problem, participants, method, results, conclusions – but in the client's language and in the client's order: first the bottom line (the main conclusion, as you wrote it), then what was examined and among whom (the question, the sample, when), what was found – for each hypothesis one sentence without symbols: what the direction is, whether it is significant, and what the effect size is in words (small/medium/large according to the conventions in the section "The effect size in words"), with the statistics in parentheses for those who want them – what it means (the practical implication you wrote for each hypothesis), limitations and recommendations. Two APA commandments are kept here too: the opening of the discussion – a clear statement on every hypothesis, without blurring "not supported" (section 3.8), and the instruction not to hide uncomfortable results (section 3.7). And one commandment of section 3.3 is deliberately dropped: the academic abstract "does not evaluate"; the executive summary does say what to do – in the recommendations section, and only there.

The results chapter

The structure – as in the first generator, following APA (2020, section 3.7): an opening paragraph (sample, design, missing data, the significance level and the confidence level – "one level throughout the paper"), and then the hypotheses in the order primary → secondary → exploratory, each with a report sentence: a description of the finding in words, the statistic with its degrees of freedom, exact p, the effect size, the confidence interval, and what it means for the hypothesis. The chapter also reports what was not supported – "do not hide uncomfortable results by omission". At the end – a checklist: for each hypothesis, whether the statistic, p, the effect size and the confidence interval were filled in, and whether the missing data and N were reported. Tables and figures are not built here – they come from the calculators (the first generator) or from the software; the reporting guidelines page says when to use a table and when a sentence.

How to write the numbers

The generator accepts the values as they were printed and fixes the form according to APA (2020, section 6.36): values that cannot exceed 1 – p, r, r_s, β, η², α, κ – without a zero before the decimal point (.03, not 0.03); values that can – t, F, d, M, SD, B – with a zero; exact p to three decimal places, and below one thousandth p < .001 (section 6.36); a negative value – with a minus sign (section 6.45); a confidence interval in the format 95% CI [LL, UL] (section 6.43). The statistic is written as the calculator writes it – t(114) = 2.31, F(2, 117) = 4.08, χ²(1, N = 304) = 6.38, r(118) = .24, U = 1283.00, z = −2.72 – and the parenthetical formats of Field, Pallant and Gravetter and Wallnau for each test are in Table 2 on the guidelines page. Statistical symbols are not defined (section 6.44) – but CI is: "95% CI" appears once in full in the chapter ("confidence interval").

The effect size in words

In the executive summary the effect is translated into a word according to the conventions the calculators use (all as cited in Field, 2024, and Pallant, 2020): Cohen's d – 0.2 small, 0.5 medium, 0.8 large (Cohen, 1988, as cited in Field, Section 3.7.2); r (and also r_rb) – .1 / .3 / .5 (Cohen, as cited in Pallant, Chapter 11); η² and ω² – .01 / .06 / .14 (Pallant, Chapter 17); R² – reported as the percentage of explained variance, without a label. Other measures (φ, Cramér's V, OR, κ, ε², H) are reported as a number – the threshold for V depends on the degrees of freedom, and OR depends on the base rate, so there is no single word for them. Field warns (section 3.7.3) that the thresholds "were never meant to be rules" – the word is a shortcut for the client, and the number stays in parentheses.

Common mistakes

1. An abstract that evaluates. "The study proves unequivocally" – the abstract reports (3.3); evaluation is left to the discussion, and recommendation – to the executive summary.
2. An abstract without numbers. APA asks for effect sizes and confidence intervals in the abstract's results, not just "a significant relationship was found".
3. Blurring a hypothesis that was not supported. In the executive summary too: "no difference was found" is a finding, not a failure; the order of APA 3.8 – support or nonsupport for every hypothesis, explicitly.
4. Writing "significant" without an effect. In a large sample everything is significant; the client needs to know whether the effect is small or large – hence the word and the number.
5. Causality from a cross-sectional survey. "The training increased commitment" is possible only in an experiment with random assignment; in a survey and a quasi-experiment – "is associated with" (Field, 2024, section 1.6.2; see correlational or experimental – what can be inferred). The generator adds a fixed sentence to the limitations section of the executive summary according to the research design you chose – and in a correlational or quasi-experimental design the sentence says that the findings describe associations and differences, not cause and effect. The wording of the finding itself is yours – in the "finding in words" field write "higher" and "associated", not "caused".
6. Over 250 words. The generator counts; whatever exceeds is shortened, usually in the method.
7. Repeating the title in the first sentence. The opening sentence is the problem under investigation, not "This study examined…" (3.3: "do not waste space by repeating the title").
8. Different confidence levels for different hypotheses. Choose once in the form – for the whole chapter.

References

American Psychological Association. (2020). Publication manual of the American Psychological Association (7th ed.). https://doi.org/10.1037/0000165-000

Field, A. (2024). Discovering statistics using IBM SPSS statistics (6th ed.). SAGE.

Pallant, J. (2020). SPSS survival manual: A step by step guide to data analysis using IBM SPSS (7th ed.). Routledge.

Cohen (1988) – as cited in Field (2024) and in Pallant (2020). Per the site rule, the reference list includes only sources that were read in the original.