Precision agriculture MOOC wins over Palestinian learners, but confidence lags satisfaction

When researchers at Al-Istiqlāl University in Jericho opened a free online course on precision agriculture to learners across Palestinian higher education, they were testing a bold proposition: that a carefully structured massive open online course, or MOOC, could deliver specialised training in sensor-based, geospatial and data-driven farming technologies to a region where such instruction has long been scarce. A new cross-sectional evaluation published in Frontiers in Education suggests the proposition holds up, at least at the level of learner perception. Of 69 learners invited to evaluate the course, 62 returned valid questionnaires, a striking 89.9 percent response rate, and their answers paint a coherent picture of a course that learners found enjoyable, well designed and confidence-building, with one intriguing exception.

The course itself was developed within the BENEFIT Erasmus+ project, a consortium of European and Palestinian universities that ran from 2019 to 2023 to build capacity in precision agriculture. The MOOC combined two thematic modules, an introduction to precision agriculture and a module on remote sensing and data processing, each following a uniform weekly structure: a five-minute introductory video presenting objectives, a curated set of video lectures, PDFs, web links and open-access research papers, discussion forums and a formative quiz. The course was uploaded to the Al-Istiqlāl University Moodle platform in October 2023 and opened for enrolment in December 2023.

What makes this course methodologically interesting is its instructional architecture, which integrates three complementary frameworks. The six-step ASSURE model provided the overall development scaffold, guiding designers to analyse learners, state objectives, select methods and media, utilise those materials, require learner participation and evaluate and revise. The TPACK framework, which maps the integration of technological, pedagogical and content knowledge, guided how technology, teaching craft and subject matter were woven together. Finally, the Kirkpatrick model was applied within the final ASSURE phase to assess learner reactions, the first of its four evaluation levels. The evaluation reported here is confined to that reaction level and to self-reported confidence; it does not include objective knowledge testing, which would correspond to Kirkpatrick Level 2.

To measure learner perceptions, the team adapted the Instructional Materials Motivation Survey, an instrument that operationalises the four components of Keller’s ARCS motivational model: Attention, Relevance, Confidence and Satisfaction. A panel of five subject-matter experts, three precision-agriculture specialists and two educational technologists, reviewed each candidate item for relevance, clarity and cultural appropriateness for a Palestinian higher-education audience, rating relevance on a four-point scale and computing item-level content-validity indices. Items scoring below 0.80 were revised or excluded. The resulting course-evaluation section comprised eight satisfaction items and nine confidence items, rated on a five-point Likert scale anchored at Strongly Disagree (0) and Strongly Agree (4). Five negatively worded items were reverse-scored before composite calculation, so that higher scores uniformly indicate more positive outcomes. The instrument was piloted with eight learners who had completed Module 1, who were then excluded from the main survey, and completion took roughly eight minutes.

The headline statistical results are unambiguous. Learners reported positive overall satisfaction, with a composite mean of 2.74 on the 0-to-4 scale (standard deviation 0.57), and positive overall confidence, with a composite mean of 2.54 (standard deviation 0.40). Both composites sat significantly above the scale midpoint of 2.0, corresponding to Neutral, with p-values below 0.001 and standardised effect sizes exceeding 1.2, large effects by conventional benchmarks. Agreement was strongest for perceived MOOC design quality, where 82.3 percent agreed or strongly agreed that it was a pleasure to work on such a well-designed course, followed by course enjoyment, confidence in mastering the content, and confidence in passing an examination, each at 75.8 percent agreement. The single negatively worded satisfaction item, on whether repetition caused boredom, was rejected by 51.6 percent of respondents, indicating that repetition was not perceived as a source of disengagement.

Within the confidence items, the researchers distinguished, on theoretical grounds rather than through factor analysis, a five-item Positive Confidence subscale and a four-item Perceived Difficulty or Barrier subscale. The Positive Confidence subscale yielded a mean of 2.67 (Cronbach’s alpha 0.814), while the reverse-scored Difficulty subscale yielded 2.38 (alpha 0.769). Three of the four difficulty items were rejected by majorities: 58.1 percent disagreed that the exercises were too difficult, 59.7 percent disagreed that they could not understand quite a bit of the material, and 53.2 percent disagreed that the course was more difficult than they would like. The exception was the item on slide information density, asking whether many slides carried so much information that it was hard to remember key points. This item did not differ significantly from the neutral midpoint (p = 0.742), with 51.6 percent of respondents neutral and 24.2 percent agreeing. The authors are careful to frame this as an area that may warrant further, exploratory review rather than a confirmed design weakness, given the non-significant result and the modal neutral response.

Perhaps the most striking finding is the strong positive correlation between satisfaction and confidence: Pearson’s r = 0.746, with a 95 percent confidence interval from 0.610 to 0.839, and p < 0.001. A non-parametric Spearman sensitivity check yielded a consistent result (ρ = 0.757). Learners who reported higher satisfaction with the MOOC also reported higher confidence in their learning. However, because both constructs were measured concurrently by self-report within the same instrument and at the same measurement occasion, this association cannot establish the direction of the relationship, whether one construct influences the other, or whether part of the association reflects shared measurement context rather than a substantive link between the two constructs.

A paired-samples t-test confirmed that satisfaction was significantly higher than confidence, with a mean difference of 0.20 points, p < 0.001, and a standardised paired effect size of dz = 0.53, a medium effect. The researchers interpret this satisfaction-confidence gap as theoretically consistent with the inherent technicality of precision agriculture, a field spanning remote sensing, drone operation, variable-rate technology and precision irrigation, each involving non-trivial learning curves. A confidence composite above the midpoint but below satisfaction should be read not as instructional failure but as a pattern consistent with appropriate calibration: learners were engaged and satisfied while the confidence data suggest that mastery of complex content may require sustained effort beyond a single online course.

The demographic profile of respondents adds interpretive texture. Female learners constituted the majority of the sample at 58.1 percent, a noteworthy distribution given the historically male-dominated character of agricultural science and precision-agriculture technology fields. The largest educational group held bachelor’s degrees (53.2 percent), followed by master’s (19.4 percent), doctoral (14.5 percent) and intermediate diploma qualifications (12.9 percent). The age distribution skewed young, with 64.5 percent under 25 years of age, and 67.7 percent were full-time students while 24.2 percent were full-time employees. Perhaps most tellingly, 66.1 percent of respondents had no prior MOOC experience, suggesting that the BENEFIT course served as a point of first contact with open online learning for the majority of participants. The most frequently cited enrolment motivations were improving general knowledge (40.3 percent) and professional development (30.6 percent).

The authors acknowledge several limitations that temper the strength of their conclusions. The cross-sectional, self-report design, with no control group, no pre-test and no objective performance measure, precludes causal inference and limits claims to perceptual outcomes at Kirkpatrick’s reaction level. The sample of 62 is geographically and institutionally homogeneous, drawn from a single programme, and while the response rate was high, no data were available to compare respondents with the seven non-respondents, so non-response bias cannot be excluded. No a priori power calculation was performed; sensitivity analysis indicates the sample was sufficient to detect only medium or larger effects. The two-part confidence structure was specified conceptually rather than confirmed by factor analysis, and the reliability of the complete eight-item satisfaction composite, including the reverse-scored boredom item, remains to be established. Because satisfaction and confidence were both collected by self-report within the same questionnaire, common-method variance may inflate the observed correlation to an extent that cannot be separately quantified with the present data.

Subject of Research: Learner satisfaction and confidence in a precision agriculture MOOC for Palestinian higher education

Article Title: Evaluating learner satisfaction and confidence in a precision agriculture MOOC: a cross-sectional study in Palestinian higher education

Article References: Khalilia, W., Horoub, H., Iwidat, H., Abdelhafid, F., Židek, R., Belej, Ľ., & Palková, Z. (2026). Evaluating learner satisfaction and confidence in a precision agriculture MOOC: a cross-sectional study in Palestinian higher education. Frontiers in Education, 11, Article 1954745. https://doi.org/10.3389/feduc.2026.1954745

Image Credits: AI Generated

DOI: 10.3389/feduc.2026.1954745

Keywords: MOOC, precision agriculture, Palestinian higher education, learner satisfaction, learner confidence, instructional design, ASSURE model, TPACK, Kirkpatrick model, Erasmus+, BENEFIT project, Frontiers in Education