diff --git a/app/services/admin_data_loader.rb b/app/services/admin_data_loader.rb
index 99cfa00e..aba0ad06 100644
--- a/app/services/admin_data_loader.rb
+++ b/app/services/admin_data_loader.rb
@@ -4,6 +4,7 @@ require 'csv'
 
 class AdminDataLoader
   def self.load_data(filepath:)
+    admin_data_values = []
     CSV.parse(File.read(filepath), headers: true) do |row|
       score = likert_score(row:)
       unless valid_likert_score(likert_score: score)
@@ -12,8 +13,10 @@ class AdminDataLoader
         admin data item #{admin_data_item(row:)} "
         next
       end
-      create_admin_data_value(row:, score:)
+      admin_data_values << create_admin_data_value(row:, score:)
     end
+
+    AdminDataValue.import(admin_data_values.flatten.compact, on_duplicate_key_update: :all)
   end
 
   private
@@ -24,7 +27,8 @@ class AdminDataLoader
 
   def self.likert_score(row:)
     likert_score = (row['LikertScore'] || row['Likert Score'] || row['Likert_Score']).to_f
-    round_up_to_one(likert_score:)
+    likert_score = round_up_to_one(likert_score:)
+    round_down_to_five(likert_score:)
   end
 
   def self.round_up_to_one(likert_score:)
@@ -32,6 +36,11 @@ class AdminDataLoader
     likert_score
   end
 
+  def self.round_down_to_five(likert_score:)
+    likert_score = 5 if likert_score > 5
+    likert_score
+  end
+
   def self.ay(row:)
     row['Academic Year'] || row['AcademicYear']
   end
@@ -45,10 +54,13 @@ class AdminDataLoader
   end
 
   def self.create_admin_data_value(row:, score:)
-    AdminDataValue.create!(likert_score: score,
-                           academic_year: AcademicYear.find_by_range(ay(row:)),
-                           school: School.find_by_dese_id(dese_id(row:).to_i),
-                           admin_data_item: AdminDataItem.find_by_admin_data_item_id(admin_data_item(row:)))
+    admin_data_value = AdminDataValue.find_or_initialize_by(school: School.find_by_dese_id(dese_id(row:).to_i),
+                                                            academic_year: AcademicYear.find_by_range(ay(row:)),
+                                                            admin_data_item: AdminDataItem.find_by_admin_data_item_id(admin_data_item(row:)))
+    return nil if admin_data_value.likert_score == score
+
+    admin_data_value.likert_score = score
+    admin_data_value
   end
 
   private_class_method :valid_likert_score
diff --git a/lib/tasks/data.rake b/lib/tasks/data.rake
index 2a2cb1f6..15198bfe 100644
--- a/lib/tasks/data.rake
+++ b/lib/tasks/data.rake
@@ -1,11 +1,11 @@
-require "csv"
+require 'csv'
 
 namespace :data do
-  desc "load survey responses"
+  desc 'load survey responses'
   task load_survey_responses: :environment do
     survey_item_response_count = SurveyItemResponse.count
     student_count = Student.count
-    path = "/data/survey_responses/clean/"
+    path = '/data/survey_responses/clean/'
     Sftp::Directory.open(path:) do |file|
       SurveyResponsesDataLoader.from_file(file:)
     end
@@ -16,30 +16,30 @@ namespace :data do
     end
     puts "=====================> Completed loading #{Student.count - student_count} students. #{Student.count} total students"
 
-    puts "Resetting race scores"
+    puts 'Resetting race scores'
     RaceScoreLoader.reset(fast_processing: false)
     puts "=====================> Completed loading #{RaceScore.count} race scores"
 
     Rails.cache.clear
   end
 
-  desc "seed only lowell"
+  desc 'seed only lowell'
   task seed_only_lowell: :environment do
     seeder = Seeder.new rules: [Rule::SeedOnlyLowell]
 
-    seeder.seed_academic_years "2016-17", "2017-18", "2018-19", "2019-20", "2020-21", "2021-22", "2022-23"
-    seeder.seed_districts_and_schools Rails.root.join("data", "master_list_of_schools_and_districts.csv")
-    seeder.seed_surveys Rails.root.join("data", "master_list_of_schools_and_districts.csv")
-    seeder.seed_respondents Rails.root.join("data", "master_list_of_schools_and_districts.csv")
-    seeder.seed_sqm_framework Rails.root.join("data", "sqm_framework.csv")
-    seeder.seed_demographics Rails.root.join("data", "demographics.csv")
+    seeder.seed_academic_years '2016-17', '2017-18', '2018-19', '2019-20', '2020-21', '2021-22', '2022-23'
+    seeder.seed_districts_and_schools Rails.root.join('data', 'master_list_of_schools_and_districts.csv')
+    seeder.seed_surveys Rails.root.join('data', 'master_list_of_schools_and_districts.csv')
+    seeder.seed_respondents Rails.root.join('data', 'master_list_of_schools_and_districts.csv')
+    seeder.seed_sqm_framework Rails.root.join('data', 'sqm_framework.csv')
+    seeder.seed_demographics Rails.root.join('data', 'demographics.csv')
   end
 
-  desc "load survey responses for lowell schools"
+  desc 'load survey responses for lowell schools'
   task load_survey_responses_for_lowell: :environment do
     survey_item_response_count = SurveyItemResponse.count
     student_count = Student.count
-    path = "/data/survey_responses/clean/"
+    path = '/data/survey_responses/clean/'
     Sftp::Directory.open(path:) do |file|
       SurveyResponsesDataLoader.from_file(file:)
     end
@@ -50,44 +50,44 @@ namespace :data do
     end
     puts "=====================> Completed loading #{Student.count - student_count} students. #{Student.count} total students"
 
-    puts "Resetting race scores"
+    puts 'Resetting race scores'
     RaceScoreLoader.reset(fast_processing: false)
     puts "=====================> Completed loading #{RaceScore.count} race scores"
 
     Rails.cache.clear
   end
 
-  desc "load students for lowell"
+  desc 'load students for lowell'
   task load_students_for_lowell: :environment do
     SurveyItemResponse.update_all(student_id: nil)
     StudentRace.delete_all
     Student.delete_all
 
-    Sftp::Directory.open(path: "/data/survey_responses/clean/") do |file|
+    Sftp::Directory.open(path: '/data/survey_responses/clean/') do |file|
       StudentLoader.from_file(file:, rules: [Rule::SkipNonLowellSchools])
     end
     puts "=====================> Completed loading #{Student.count - student_count} students. #{Student.count} total students"
 
-    puts "Resetting race scores"
+    puts 'Resetting race scores'
     RaceScoreLoader.reset(fast_processing: false)
     puts "=====================> Completed loading #{RaceScore.count} survey responses"
 
     Rails.cache.clear
   end
 
-  desc "delete non-lowell schools and districts"
+  desc 'delete non-lowell schools and districts'
   task delete_non_lowell: :environment do
-    schools = School.all.reject { |s| s.district.name == "Lowell" }
+    schools = School.all.reject { |s| s.district.name == 'Lowell' }
     ResponseRate.where(school: schools).delete_all
     Respondent.where(school: schools).delete_all
     Survey.where(school: schools).delete_all
     schools.each { |school| school.delete }
-    districts = District.all.reject { |district| district.name == "Lowell" }
+    districts = District.all.reject { |district| district.name == 'Lowell' }
     districts.each { |district| district.delete }
   end
 
   task load_survey_responses_21_22: :environment do
-    Dir.glob(Rails.root.join("data", "survey_responses", "*2021-22*.csv")).each do |filepath|
+    Dir.glob(Rails.root.join('data', 'survey_responses', '*2021-22*.csv')).each do |filepath|
       puts "=====================> Loading data from csv at path: #{filepath}"
       SurveyResponsesDataLoader.load_data filepath:
     end
@@ -95,7 +95,7 @@ namespace :data do
   end
 
   task load_survey_responses_20_21: :environment do
-    Dir.glob(Rails.root.join("data", "survey_responses", "*2020-21*.csv")).each do |filepath|
+    Dir.glob(Rails.root.join('data', 'survey_responses', '*2020-21*.csv')).each do |filepath|
       puts "=====================> Loading data from csv at path: #{filepath}"
       SurveyResponsesDataLoader.load_data filepath:
     end
@@ -103,7 +103,7 @@ namespace :data do
   end
 
   task load_survey_responses_19_20: :environment do
-    Dir.glob(Rails.root.join("data", "survey_responses", "*2019-20*.csv")).each do |filepath|
+    Dir.glob(Rails.root.join('data', 'survey_responses', '*2019-20*.csv')).each do |filepath|
       puts "=====================> Loading data from csv at path: #{filepath}"
       SurveyResponsesDataLoader.load_data filepath:
     end
@@ -111,7 +111,7 @@ namespace :data do
   end
 
   task load_survey_responses_18_19: :environment do
-    Dir.glob(Rails.root.join("data", "survey_responses", "*2018-19*.csv")).each do |filepath|
+    Dir.glob(Rails.root.join('data', 'survey_responses', '*2018-19*.csv')).each do |filepath|
       puts "=====================> Loading data from csv at path: #{filepath}"
       SurveyResponsesDataLoader.load_data filepath:
     end
@@ -119,7 +119,7 @@ namespace :data do
   end
 
   task load_survey_responses_17_18: :environment do
-    Dir.glob(Rails.root.join("data", "survey_responses", "*2017-18*.csv")).each do |filepath|
+    Dir.glob(Rails.root.join('data', 'survey_responses', '*2017-18*.csv')).each do |filepath|
       puts "=====================> Loading data from csv at path: #{filepath}"
       SurveyResponsesDataLoader.load_data filepath:
     end
@@ -127,86 +127,86 @@ namespace :data do
   end
 
   task load_survey_responses_16_17: :environment do
-    Dir.glob(Rails.root.join("data", "survey_responses", "*2016-17*.csv")).each do |filepath|
+    Dir.glob(Rails.root.join('data', 'survey_responses', '*2016-17*.csv')).each do |filepath|
       puts "=====================> Loading data from csv at path: #{filepath}"
       SurveyResponsesDataLoader.load_data filepath:
     end
     puts "=====================> Completed loading #{SurveyItemResponse.count} survey responses"
   end
 
-  desc "reset response rate values"
+  desc 'reset response rate values'
   task reset_response_rates: :environment do
-    puts "Resetting response rates"
+    puts 'Resetting response rates'
     ResponseRateLoader.reset
     Rails.cache.clear
     puts "=====================> Completed loading #{ResponseRate.count} survey responses"
   end
 
-  desc "reset race score calculations"
+  desc 'reset race score calculations'
   task reset_race_scores: :environment do
-    puts "Resetting race scores"
+    puts 'Resetting race scores'
     RaceScoreLoader.reset(fast_processing: false)
     Rails.cache.clear
     puts "=====================> Completed loading #{RaceScore.count} survey responses"
   end
 
-  desc "load admin_data"
+  desc 'load admin_data'
   task load_admin_data: :environment do
-    AdminDataValue.delete_all
-    Dir.glob(Rails.root.join("data", "admin_data", "dese", "*.csv")).each do |filepath|
+    original_count = AdminDataValue.count
+    Dir.glob(Rails.root.join('data', 'admin_data', 'dese', '*.csv')).each do |filepath|
       puts "=====================> Loading data from csv at path: #{filepath}"
       Dese::Loader.load_data filepath:
     end
-    puts "=====================> Completed loading #{AdminDataValue.count} survey responses"
+    puts "=====================> Completed loading #{AdminDataValue.count - original_count} admin data values"
   end
 
-  desc "load students"
+  desc 'load students'
   task load_students: :environment do
     SurveyItemResponse.update_all(student_id: nil)
     StudentRace.delete_all
     Student.delete_all
-    Dir.glob(Rails.root.join("data", "survey_responses", "*student*.csv")).each do |file|
+    Dir.glob(Rails.root.join('data', 'survey_responses', '*student*.csv')).each do |file|
       puts "=====================> Loading student data from csv at path: #{file}"
       StudentLoader.load_data filepath: file
     end
     puts "=====================> Completed loading #{Student.count} students"
 
-    puts "Resetting race scores"
+    puts 'Resetting race scores'
     RaceScoreLoader.reset(fast_processing: false)
     puts "=====================> Completed loading #{RaceScore.count} survey responses"
 
     Rails.cache.clear
   end
 
-  desc "reset all cache counters"
+  desc 'reset all cache counters'
   task reset_cache_counters: :environment do
-    puts "=====================> Resetting Category counters"
+    puts '=====================> Resetting Category counters'
     Category.all.each do |category|
       Category.reset_counters(category.id, :subcategories)
     end
-    puts "=====================> Resetting Subcategory counters"
+    puts '=====================> Resetting Subcategory counters'
     Subcategory.all.each do |subcategory|
       Subcategory.reset_counters(subcategory.id, :measures)
     end
-    puts "=====================> Resetting Measure counters"
+    puts '=====================> Resetting Measure counters'
     Measure.all.each do |measure|
       Measure.reset_counters(measure.id, :scales)
     end
-    puts "=====================> Resetting Scale counters"
+    puts '=====================> Resetting Scale counters'
     Scale.all.each do |scale|
       Scale.reset_counters(scale.id, :survey_items)
     end
-    puts "=====================> Resetting SurveyItem counters"
+    puts '=====================> Resetting SurveyItem counters'
     SurveyItem.all.each do |survey_item|
       SurveyItem.reset_counters(survey_item.id, :survey_item_responses)
     end
   end
 
-  desc "scrape dese site for admin data"
+  desc 'scrape dese site for admin data'
   task scrape_all: :environment do
-    puts "scraping data from dese"
+    puts 'scraping data from dese'
     scrapers = [Dese::OneAOne, Dese::OneAThree, Dese::TwoAOne, Dese::TwoCOne, Dese::ThreeAOne, Dese::ThreeATwo,
-      Dese::ThreeBOne, Dese::ThreeBTwo, Dese::FourAOne, Dese::FourBTwo, Dese::FourDOne, Dese::FiveCOne, Dese::FiveDTwo]
+                Dese::ThreeBOne, Dese::ThreeBTwo, Dese::FourAOne, Dese::FourBTwo, Dese::FourDOne, Dese::FiveCOne, Dese::FiveDTwo]
     scrapers.each do |scraper|
       scraper.new.run_all
     end
diff --git a/spec/fixtures/secondary_sample_admin_data.csv b/spec/fixtures/secondary_sample_admin_data.csv
new file mode 100644
index 00000000..8cdc3c2d
--- /dev/null
+++ b/spec/fixtures/secondary_sample_admin_data.csv
@@ -0,0 +1,13 @@
+District,School,DESE ID,Category,Item ID,NonLikert Title,NL_Value,LikertScore,Benchmark,Data Type,Academic Year
+Attleboro,Attleboro High School,160505,2-C-i,a-vale-i1,Chronic absence rate,19.7,1,10,%,2018-19
+Milford,Woodland Elementary School,1850090,2-C-i,a-vale-i1,Chronic absence rate,6.8,1,10,%,2018-19
+Revere,Beachmont Elementary School,2480013,2-C-i,a-vale-i1,Chronic absence rate,4.2,1,10,%,2018-19
+Winchester,Winchester High School,3440505,2-C-i,a-vale-i1,Chronic absence rate,7.2,1,10,%,2018-19
+Attleboro,Attleboro High School,160505,3-A-i,a-reso-i1,Average class size,20.6,2,20,,2018-19
+Milford,Woodland Elementary School,1850090,3-A-i,a-reso-i1,Average class size,22.5,2,20,,2018-19
+Revere,Beachmont Elementary School,2480013,3-A-i,a-reso-i1,Average class size,17,2,20,,2018-19
+Winchester,Winchester High School,3440505,3-A-i,a-reso-i1,Average class size,17,2,20,,2018-19
+Attleboro,Attleboro High School,160505,3-A-ii,a-sust-i3,Student to instructional support staff ratio,15.28896673,3,43.4,,2018-19
+Milford,Woodland Elementary School,1850090,3-A-ii,a-sust-i3,Student to instructional support staff ratio,22.85714286,3,43.4,,2018-19
+Revere,Beachmont Elementary School,2480013,3-A-ii,a-sust-i3,Student to instructional support staff ratio,38,3,43.4,,2018-19
+Winchester,Winchester High School,3440505,3-A-ii,a-sust-i3,Student to instructional support staff ratio,135.9,3,43.4,,2018-19
diff --git a/spec/services/admin_data_loader_spec.rb b/spec/services/admin_data_loader_spec.rb
index 09f10a26..b7ca5c3d 100644
--- a/spec/services/admin_data_loader_spec.rb
+++ b/spec/services/admin_data_loader_spec.rb
@@ -2,6 +2,7 @@ require 'rails_helper'
 
 describe AdminDataLoader do
   let(:path_to_admin_data) { Rails.root.join('spec', 'fixtures', 'sample_admin_data.csv') }
+  let(:path_to_secondary_admin_data) { Rails.root.join('spec', 'fixtures', 'secondary_sample_admin_data.csv') }
   let(:ay_2018_19) { AcademicYear.find_by_range '2018-19' }
   let(:attleboro) { School.find_by_dese_id 160_505 }
   let(:winchester) { School.find_by_dese_id 3_440_505 }
@@ -29,7 +30,7 @@ describe AdminDataLoader do
 
       # it 'assigns the school to the admin data value' do
       expect(AdminDataValue.first.school).to eq attleboro
-      expect(AdminDataValue.last.school).to eq beachmont
+      expect(AdminDataValue.last.school).to eq winchester
       # end
 
       # it 'links the admin data value to the correct admin data item' do
@@ -38,7 +39,7 @@ describe AdminDataLoader do
       # end
 
       # it 'loads all the admin data values in the target csv file' do
-      expect(AdminDataValue.count).to eq 10
+      expect(AdminDataValue.count).to eq 11
       # end
 
       # it 'captures the likert score ' do
@@ -46,6 +47,8 @@ describe AdminDataLoader do
                                     admin_data_item: chronic_absense_rate).likert_score).to eq 3.03
       expect(AdminDataValue.find_by(school: beachmont, academic_year: ay_2018_19,
                                     admin_data_item: student_to_instructor_ratio).likert_score).to eq 3.5
+      expect(AdminDataValue.find_by(school: winchester, academic_year: ay_2018_19,
+                                    admin_data_item: student_to_instructor_ratio).likert_score).to eq 5
       # end
 
       # it 'rounds up any likert_scores between 0 and 1 (non-inclusive) to 1' do
@@ -55,16 +58,27 @@ describe AdminDataLoader do
       # it 'rejects importing rows with a value of 0' do
       expect(AdminDataValue.where(school: attleboro, academic_year: ay_2018_19,
                                   admin_data_item: AdminDataItem.find_by_admin_data_item_id('a-reso-i1'))).not_to exist
-      expect(AdminDataValue.where(school: winchester, academic_year: ay_2018_19,
-                                  admin_data_item: AdminDataItem.find_by_admin_data_item_id('a-sust-i3'))).not_to exist
       # end
     end
+
+    context 'when a second file exists' do
+      before :each do
+        AdminDataLoader.load_data filepath: path_to_secondary_admin_data
+      end
+
+      it 'updates likert scores to match the new file' do
+        expect(AdminDataValue.find_by(school: attleboro, academic_year: ay_2018_19,
+                                      admin_data_item: chronic_absense_rate).likert_score).to eq 1
+        expect(AdminDataValue.find_by(school: beachmont, academic_year: ay_2018_19,
+                                      admin_data_item: student_to_instructor_ratio).likert_score).to eq 3
+      end
+    end
   end
 
   describe 'output to console' do
     it 'outputs a messsage saying a value has been rejected' do
-      output = capture_stdout { AdminDataLoader.load_data filepath: path_to_admin_data }.gsub("\n", '')
-      expect(output).to eq 'Invalid score: 0.0        for school: Attleboro High School        admin data item a-reso-i1 Invalid score: 100.0        for school: Winchester High School        admin data item a-sust-i3 '
+      output = capture_stdout { AdminDataLoader.load_data filepath: path_to_admin_data }.delete("\n")
+      expect(output).to eq 'Invalid score: 0.0        for school: Attleboro High School        admin data item a-reso-i1 '
     end
   end
 end