[Context] Online Recruitment and Selection (R&S) processes are often the first point of contact between early-career software engineers and the tech industry. Yet many candidates experience these processes as opaque, inefficient, or even discouraging. While prior research has extensively documented the flaws and biases in online tech hiring, little is known about the practices that create positive candidate experiences. [Objective & Method] This paper explores such practices, referred to as Constructive Patterns (CPs), from the perspective of early-career software engineers. Guided by Applicant Attribution-Reaction Theory, we conducted 22 semi-structured interviews in which participants collectively described over 470 online R&S experiences. [Results] Through thematic analysis, we identified 22 CPs that reflect positive practices such as comprehensive and transparent job advertisements (CP01), specific and developmental feedback (CP03), humanized and respectful interaction (CP06), and framing the process as a two-way street (CP18). [Conclusion] Our findings extend the conversation on tech hiring beyond diagnosing dysfunctions toward designing for human-centered and growth-oriented candidate experiences. The resulting catalog of CPs provides a concrete and empirically grounded resource for organizations seeking to attract and support early-career software engineers more effectively.

翻译：[背景] 在线招聘与选拔流程通常是早期职业软件工程师与技术行业的首次接触点。然而，许多候选人认为这些流程不透明、效率低下，甚至令人沮丧。尽管先前研究已广泛记录了在线技术招聘中的缺陷与偏见，但关于创造积极候选人体验的实践却知之甚少。[目标与方法] 本文从早期职业软件工程师的视角，探讨了被称为"建设性模式"的此类实践。在申请人归因反应理论的指导下，我们进行了22次半结构化访谈，参与者共同描述了超过470次在线招聘与选拔经历。[结果] 通过主题分析，我们识别出22种建设性模式，这些模式反映了积极的实践，例如全面透明的职位广告（CP01）、具体且具有发展性的反馈（CP03）、人性化且尊重的互动（CP06），以及将流程构建为双向选择过程（CP18）。[结论] 我们的研究将技术招聘的讨论从诊断功能障碍，扩展到为以人为本和成长导向的候选人体验进行设计。由此产生的建设性模式目录为寻求更有效吸引和支持早期职业软件工程师的组织提供了具体且基于实证的资源。