There　is　strong　heterogeneity　and　correlation　in　the　connection　mode　choice　of　high-speed　railway　passenger　hubs.　In-depth　analysis　of　heterogeneity　and　correlation　can　provide　support　for　multi-mode　connection　capacity　scheduling　of　high-speed　railway　passenger　hubs.　The　RP　/　SP　questionnaire　survey　is　used　to　obtain　data　on　passengers　＇　individual　socioeconomic　attributes,　hub　transfer　characteristics　and　connection　mode　choice　in　assumed　travel　scenarios.　A　hybrid　method　is　proposed　to　concurrently　realize　the　analysis　of　heterogeneity　and　correlation　in　connection　mode　choice.　The　latent-class　conditional　logit　model　is　employed　to　categorize　passengers,　exploring　the　heterogeneity　among　various　categories　of　passengers.　The　Apriori　algorithm　is　utilized　to　analyze　the　correlation　between　personal　socioeconomic　attributes　or　assumed　travel　scenarios　and　passenger　choice　of　connecting　mode.　The　results　show　that　passengers　can　be　divided　into　three　categories:　self-paying　on-demand　transportation　enthusiasts,　non-self-paying　on-demand　transportation　enthusiasts　and　route　transportation　enthusiasts.　The　socioeconomic　attributes　of　individuals,　including　occupation,　educational　background　and　annual　income,　have　significant　impacts　on　the　categories.　Passenger　connection　mode　choices　among　different　categories　differ　in　individual　socioeconomic　attributes,　while　passenger　connection　mode　choices　in　the　same　category　differ　in　assumed　travel　scenarios.　The　correlation　between　personal　socioeconomic　attributes　and　assumed　travel　scenarios　may　affect　passenger　preferences.　©　2025　Chinese　Society　of　Civil　Engineering.　All　rights　reserved.